Wi-Fi: Connectivity Backbone for Devices Tools Take a High

Transcription

The magazine of record for the embedded computing industry
July 2012
www.rtcmagazine.com
Wi-Fi: Connectivity
Backbone for Devices
Tools Take a High-Level
Look at Code
An RTC Group Publication
Have We Yet Pulled All the
Power out of Multicore?
OpenCL
Unleashes
the Power
of Parallel Coprocessing
58
Rugged Router Runs Cisco IOS Software
60
Ultra Low Power ARM-Based Embedded Computer
Designed for 3.5” to 12” HMI
TABLEOF CONTENTS
65
3U OpenVPX SBCs Bring 10 Gig Ethernet and PCI
Express 3.0
VOLUME 21, ISSUE 7
Departments
6
8
Editorial
The “ASIC Quandary”—How Much Can
We Integrate, When and Why?
Industry Insider
Latest Developments in the Embedded
Marketplace
Technology in Context
TECHNOLOGY DEPLOYED
The Expanding Roles of Non-Volatile
Memory
Code Requirements and Verification
16
The Expanding Role of Non-Volatile
Memory in High-Performance
Embedded Architecture
Adrian Proctor, Viking Technology
Form Factor Forum
TECHNOLOGY IN SYSTEMS
10Small
The Graying of Embedded
Developing Hybrid Code Using OpenCL
Products & Technology
Programming: Parallel
Embedded Technology Used by
58Newest
Processing Made Faster and
26 OpenCL
Industry Leaders
Easier than Ever
EDITOR’S REPORT
Realizing the Potential of Multicore
Processors
12
Retooling Applications to Ride on
Multiple Cores
Tom Williams
44Requirements Engineering Today
of the Passenger’s Seat:
48Out
Requirements Traceability to Drive
the Software Development Process
Marcia Stinson, Visure
Jared Fry, LDRA Technology
Code Analysis with
Visualization
54Transforming
Paul Anderson, GrammaTech
Todd Roberts, AMD
Embedded Hybrid Code
32 Developing
Using OpenCL
Mark Benson, Logic PD
Computing with AMD
38 Parallel
Fusion-Based Computer-onModules
John Dockstader, congatec
Digital Subscriptions Avaliable at http://rtcmagazine.com/home/subscribe.php
RTC MAGAZINE JULY 2012
3
JULY 2012
Publisher
PRESIDENT
John Reardon, [email protected]
Editorial
EDITOR-IN-CHIEF
Tom Williams, [email protected]
CONTRIBUTING EDITORS
Colin McCracken and Paul Rosenfeld
MANAGING EDITOR/ASSOCIATE PUBLISHER
Sandra Sillion, [email protected]
COPY EDITOR
Rochelle Cohn
Art/Production
ART DIRECTOR
Kirsten Wyatt, [email protected]
GRAPHIC DESIGNER
Michael Farina, [email protected]
LEAD WEB DEVELOPER
Justin Herter, [email protected]
Advertising/Web Advertising
WESTERN REGIONAL ADVERTISING MANAGER
Stacy Mannik, [email protected]
(949) 226-2024
MIDWEST REGIONAL AND INTERNATIONAL ADVERTISING MANAGER
Mark Dunaway, [email protected]
(949) 226-2023
EASTERN REGIONAL ADVERTISING MANAGER
Shandi Ricciotti, [email protected]
(949) 573-7660
Billing
Cindy Muir, [email protected]
(949) 226-2021
To Contact RTC magazine:
HOME OFFICE
The RTC Group, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673
Phone: (949) 226-2000 Fax: (949) 226-2050, www.rtcgroup.com
Editorial Office
Tom Williams, Editor-in-Chief
1669 Nelson Road, No. 2, Scotts Valley, CA 95066
Phone: (831) 335-1509
Published by The RTC Group
Copyright 2012, The RTC Group. Printed in the United States. All rights reserved. All related
graphics are trademarks of The RTC Group. All other brand and product names are the property
of their holders.
4
Untitled-5 1
JULY 2012 RTC MAGAZINE
6/4/12 2:04 PM
Higher Performance . Lower Power . Small Form Factor
Small SuperServers® and Motherboards Optimized for Embedded Appliances
X9SPV-F/LN4F
mini-ITX (6.7” x 6.7”)
t Mobile, 3rd generation Intel® Core™ i7/i5/i3 processors
and Intel® QM77 Express Chipset
t 2x / 4x Gigabit LAN Ports, 10x USB Ports
ctor
t SATA DOM (Disk on Module) Power connector
ptions
t Dual/Quad core 17W, 25W and 35W TDP options
t Up to 16GB DDR3 SO-DIMMs Memory with ECC
t 1 PCI-E 3.0 x16 expansion slot
t 2 SATA 6Gb/s and 4 SATA 3Gb/s Storage interface
face with RAID
D 0/1/5/10
0
0
New
d OpenGL 3.1)
3 1))
t Integrated Graphics ( DX11, OpenCL 1.1 and
t IPMI 2.0 with dedicated LAN for Remote Management
Applications:
t
t
t
t
t
(Front View)
SMB Servers
Storage Head Node
HD Video Conferencing
Network and Security Appliance
Digital / Network Video Surveillance
(Rear View)
SYS-5017P-TF/TLN4F
Compact Short Depth System
17.2”W x 1.7”H x 9.8”L
Next Generation Family of Motherboards for Embedded Applications
N
ns
X9SCV-Q
X
X9SC
X9
SCV
S
SC
V--Q
V
Mini-ITX
C7Q67
C7Q6
C
C7
7Q6
Q 7
Q67
Micro-ATX
X9SAE
ATX
www.supermicro.com/Embedded
© Super Micro Computer, Inc. Specifications subject to change without notice.
Intel®, the Intel® logo, Xeon®, and Xeon® Inside, are trademarks or registered trademarks of Intel Corporation in the US and /or other countries.
All other brands and names are the property of their respective owners.
X9DRD-EF
X9DR
X9
9DR
DRDDRD
D EF
E ATX
E-ATX
EDITORIAL
JULY 2012
The “ASIC Quandary”—How Much
Can We Integrate, When and Why?
W
e’ve just gotten news of a market survey on graphics shipments that reports a decline in overall graphics
shipments for the year. We will not embarrass the company by naming it because while its survey did include graphics
shipped in desktops, notebooks, netbooks and PC-based commercial and embedded equipment, it did not include ARM-based tablets or, apparently, ARM-based embedded devices. It did somewhat lamely admit that tablets have changed the PC market. Duh.
If we’re talking about shipments of graphics processors,
both discrete and integrated onto the same die with a CPU, it
makes utterly no sense to ignore the vast and growing number
of devices based on ARM processors with high-performance integrated graphics. The report claims that graphics leader Nvidia
has dropped shipments—but does not acknowledge shipments of
multicore graphics processors integrated with ARM cores on that
same company’s Tegra CPU products. These processors appear
in a wide variety of Android-based tablets. And, of course, there
is no mention of the graphics integrated on the Apple-proprietary
ARM-based CPUs built into the millions of iPads that are being
sold. So what exactly is the point of all this? Go figure.
It was easier when graphics companies sold graphics processors that could be discrete or integrated into graphics add-in modules. All this integration stuff has gotten very messy. Now whether
you think you need the high-end graphics capability or not, it comes
with your Intel Core i5 or i7; it comes with your low-power Atom
CPU, and it is a featured part of your AMD G- or R-Series APU.
One of the results has been that developers really are making use of
the powerful graphics that come with the chips they need for their
designs, and more and more possibilities are opening up for the use
of such graphics in embedded consumer and industrial applications.
It turns out, however, that this development with graphics
points to a wider phenomenon. It is now possible to integrate
practically anything we want onto a single silicon die. Of course,
just because we can does not automatically make it a good idea.
And that is the core of what we might call “the ASIC quandary.”
What, exactly, are the conditions that must be met to put together
a given combination of general vs. specialized functionality as
hard-wired silicon? What amount of configurability and programmability? What mix of peripherals and on-chip memory of
what type can be brought together that make technical, market
and economic sense? And how much of it can we really take
6
Tom Williams
Editor-in-Chief
advantage of? Of course, if you’re able to write a large enough
check you can now get an ASIC that integrates practically anything, but that is not the path of the majority of the industry.
Interestingly, graphics is something we now understand
pretty well, while multicore CPUs are something we are still
struggling to fully master (see the Editor’s Report in this issue).
Back in the 80s, a company named Silicon Graphics occupied a
huge campus in Silicon Valley. They produced a very high-end
graphics chip called the Graphics Engine that was used in their
line of advanced graphics workstations. They also developed the
graphics language that eventually became OpenGL, the most
portable and widely used graphics software today and one that
can be used for all manner of high-end GPUs.
Thus, of the mix of considerations mentioned above, market
volume acceptance of high-end graphics has made it a natural for
integration on all the latest CPUs. Multicore processors (which also
integrate advanced graphics) are certainly another case, and there
is definitely enough motivation to exploit their full potential that
these will long continue to gain in market acceptance. But what else
can we expect to eventually be massively integrated onto everyday
CPUs? Could it be DSP? Well, that appears to be already covered
by some of the massively parallel general-purpose graphics processors that are now integrated into CPUs by companies like AMD
and Nvidia. There is a language that is emerging to take advantage
of their power for numerically intensive processing that—perhaps
not accidentally similar to OpenGL—is named OpenCL.
For the present, we are also now seeing the increasing integration not of hard-wired silicon functionality but of silicon configurability and programmability. The initial Programmable System on
Chip (PSoC) from Cypress Semiconductor started as an alternative
to the dauntingly huge selection of 8-bit processors that offered a
dizzying assortment of peripheral combinations. PSoC has since
grown to include products based on 32-bit ARM devices. More recently, devices have come out of Xilinx and Altera that combine a
32-bit multicore ARM processor with its normal peripherals on the
same die with an FPGA fabric. While the market has yet to issue a
final verdict here, this indicates the direction of the ASIC quandary.
Add the ability to make choices and later, if a large enough trend
is identified, some enterprising company may issue the stuff hardwired. The ability for truly massive integration is here. The direction it takes will depend on forces both technical and economic.
INDUSTRY
INSIDER
JULY 2012
Acromag Acquires the Assets of Xembedded
On May 15, 2012, Acromag acquired the assets of Xembedded. This purchase adds the products of Xembedded
to Acromag’s portfolio, benefitting both groups by combining product lines. The company says that now customers
will have access to a complete product solution, ranging from CPU products to I/O solutions.
Xembedded products are known for quality and technology. Their focus on VME, VPX and COM Express platforms complements Acromag’s FPGA, PMC, XMC and I/O technologies. This move also provides for extending
production of the many legacy products that Xembedded, LLC was known for.
The new “Xembedded Group” will now join the Acromag “Embedded Solutions Group” and “Process Instrument
Group.” By forming this new group, Acromag hopes to provide uninterrupted service to former Xembedded customers. Along with the products, Acromag has hired all of the employees of Xembedded in order to provide the expertise
to design, manufacture and support the many products in the “Xembedded Group” portfolio. The new “Xembedded
Group” will work to continue the relationship of the many sales representatives and distributors that have sold these
products for years.
IBM and National
Instruments Collaborate on
Software Quality and Test
Software is more complex
than ever. The amount of software in today’s luxury car, for
example, is expected to increase
from almost 10 million lines of
code in 2010 to 100 million in
2030. With this increased complexity, come higher error rates.
To address these challenges,
IBM and National Instruments
have joined forces to allow system
engineering departments the ability to trace and improve development end to end, from initial requirements through final systems
test, in one single environment.
The integration will include the
leading real-time and embedded test software from National
Instruments and the Application
Lifecycle Management (ALM)
capabilities from IBM Rational.
Significantly reducing time
and cost, this integrated approach
eliminates the need to track items
manually through spreadsheets
and other tools. This offering
brings quality and test together,
with benefits that apply to engineering departments in virtually
any industry including automotive, aerospace and defense, medical and telecommunications.
8
For instance, an automotive
cruise control system must adhere
to rigorous safety test standards.
When testing, the tolerance of set
speed is violated for more than a
permitted period of time. This error can be flagged in the IBM and
National Instruments solution.
All test artifacts as well as data
needed to reproduce the error are
logged in the system automatically. Immediately, this error is
traced to software design requirements to understand the impact of
the error on customer experience.
It can then be determined if this
is a safety-critical issue and therefore assigned appropriate priority.
Near Sensor Computing
Business Unit for
Miniaturized FPGA Solutions
Interconnect Systems (ISI)
has announced that it has established a new business unit for
near sensor computing. Near sensor computing blends ISI’s advanced packaging expertise with
FPGA system design to develop
and manufacture miniaturized
solutions targeted at a wide range
of applications, including medical imaging, night vision, industrial inspection, unmanned aerial
vehicles (UAVs), helmet cameras
and surveillance.
Latest generation FPGAs
provide small, low-cost engines
that are ideal for sensor processing applications. ISI’s Near Sensor Computing modules move
the processor closer to the sensor, decreasing size, cost, latencies, bandwidth requirements and
power consumption across the
entire system. These modules are
configured to order with processor, memory and I/O resources
tailored to each customer’s specific requirements.
Current Near Sensor Computing systems in production are
the NSC-625 and the NSC-2200.
The NSC-625 is a 1-inch square
(25 mm) form factor, while the
NSC-2200 holds a 40 x 60 mm
footprint. Both utilize a miniaturized processor, memory and I/O
modules delivered pre-configured
to the specific requirements of the
customer’s near sensor computing application. These FPGA
systems perform real-time image
enhancement, sensor fusion and
object/event recognition for any
analog or digital sensor including
4 x HD IR and visible-spectrum
video. Output options include
1/10 GigE and GigE Vision,
LVDS, Camera Link, Aurora,
RS-232, programmable pins and
custom formats.
Cellular M2M Connections
in the Retail Industry
Surpassed 10 Million in 2011
The number of cellular
M2M connections in the retail
industry reached 10.3 million
worldwide in 2011, according to
a new research report from Berg
Insight. Cellular M2M technology enables devices such as POS
terminals, ATMs and ticketing
machines to be used at new locations where fixed line connectivity is unavailable or impractical.
The technology has a more transformational effect on markets
such as vending and parking,
where machine operators need to
reorganize their operations in order to benefit from the availability
of real-time information.
Berg Insight forecasts that
the number of cellular M2M connections in the global retail industry will grow at a compound
annual growth rate (CAGR) of
21.6 percent during the next six
years to reach 33.2 million connections in 2017. Shipments of
cellular M2M devices for retail
applications will at the same
time increase at a CAGR of 10.7
percent from 5.2 million units in
2011 to 9.6 million units in 2017.
“POS terminals will constitute the lion’s share of cellular
M2M connections in the retail
sector throughout the forecast
period” says Lars Kurkinen, telecom analyst, Berg Insight. “But
the penetration rate for cellular
connectivity is today actually
highest in the multi-space parking meter segment where it is well
above 30 percent” he adds. Berg
Insight expects that the vending
machine segment will present a
major opportunity for wireless
connectivity in the long term, and
the study indicates that vending
telemetry solutions will be the
fastest growing segment during
the next six years.
Paper-Thin, Non-Flammable
Battery Technology for
Consumer Devices
A new battery technology
aims at replacing conventional
lithium-ion batteries and redefining the design of smartphones,
tablets and other consumer devices by enabling thinner form
factors and better performance.
Infinite Power Solutions (IPS) has
demonstrated advanced energy
storage capacity for an organicfree, all-solid-state, thin-film battery. Using single-sided deposition, a paper-thin rechargeable
cell achieved a capacity density
of 1.25 mAh/cm2. IPS also announced it has already initiated
talks with original equipment
manufacturers for future adoption of this technology in consumer devices.
IPS has also released a
white paper documenting a path
to manufacturing this new battery technology using roll-to-roll
manufacturing and double-sided
deposition in order to achieve a
capacity density of 2.5 mAh/cm2,
which is equivalent to prismatic
lithium-ion cells (but without the
risk of thermal runaway), and at
comparable production costs.
This new solid-state battery
could be manufactured in volume
for less than $1/Wh and, when
fully packaged, provide 700 Wh/
liter—a 25 percent increase in energy density over today’s lithiumion cells.
High-capacity cells can be
arranged in a variety of configurations, including designing a
battery around the form factor
of the display of a smartphone or
tablet. For example, cells can be
safely stacked in a parallel configuration to form a 3.95V allsolid-state, thin-film battery pack
about the size and thinness of five
standard playing cards, which is
three times thinner than today’s
lithium-ion cell phone batter-
ies of equivalent capacity. Such
a battery would have a capacity
of approximately 1.4 Ah with a
maximum current of 7A and
could serve a variety of common
handheld consumer devices.
LDRA Commits to Support
All Aerospace Platforms for
Certification
LDRA, a company in standards compliance, automated
software verification, source
code analysis and test tools, has
automated the interface between
its LDRA tool suite and Atmel’s
AVR Studio. This interface helps
ensure that aerospace applications developed using Atmel’s
AVR ATmega128 are certifiable
to DO-178C Level A. Automating
the interface provides an efficient
way for developers to more easily
complete elements of DO-178C
qualification that otherwise demand a highly skilled and laborintensive process.
To achieve compliance with
the FAA’s software standard
DO-178C Level A, certification
applicants must prove that their
system and its application work as
expected not only at the sourcecode level, but also at the objectcode level. Problems determining the cause of failures in the
medical and automotive industries are also leading to the same
technique becoming increasingly
required in IEC 62304 and ISO
26262 certifications. For systems
needing such stringent compliance, the LDRA tool suite traces
requirements to the underlying
assembler code to confirm that no
errors have been introduced when
the higher-level programming
language devolves into low-level
object code.
With this interface, the
LDRA tool suite drives the test
environment to automate the
exchange of execution history
and test data using Atmel’s AVR
Studio 4.18 command-line utilities, AVRISP mkII device and
an onboard serial port. The collation of the LDRA tool suites for
both Assembler and C languages
with the AVR-GCC compiler
streamlines the interface for the
developers, enabling them to efficiently verify that the AVR-GCCgenerated object code accurately
and completely represents the application’s C source code operational requirements. Developers
can fully verify the application on
target, automating the functionality required to meet the stringent
demands of DO-178C Level A.
I/O, power and infrastructure
solutions, build technology components to HP Proactive Insight
architecture standards and interfaces. These components help to
analyze thousands of system parameters and automate processes
that consume critical data center
resources. Additionally, they provide insight into aspects of application performance and IT infrastructure optimization that had
once been nearly impossible to
capture and utilize. QLogic will
use the HP PIA qualified insignia
as visible assurance of meeting
HP’s stringent quality and performance standards.
QLogic Joins the HP
ProActive Insight
Architecture Alliance
QLogic has announced it
is a founding member of the HP
ProActive Insight Architecture
(PIA) Alliance, selected by HP
for its networking solutions developed to the HP standards and
interfaces that contribute to the
performance and reliability of HP
ProLiant Generation 8 (Gen8).
HP ProActive Insight Architecture, now available in the new
HP ProLiant Gen8 servers, transforms data center economics with
built-in intelligence that improves
performance and uptime, while
lowering operating costs. As a
founding member of the HP PIA
Alliance, QLogic will contribute
to the elimination of common
problems caused by human-totechnology interaction that lead
to system downtime and data loss.
Through embedded HP ProLiant
Gen8 server technology, HP PIA
Alliance members, along with
HP, can evaluate data and other
critical analytics to continually
optimize and improve on business results.
Alliance members, drawn
from leaders in memory, storage,
9
SMALL FORM FACTOR
FORUM
Colin McCracken & Paul Rosenfeld
The Graying of Embedded
E
mbedded computing may be heading for some new territory. And success in that territory calls for new strategies and skills. Not that embedded will ever go away, of
course. Once focused on unique application I/O and driven by
desktop-derived or purpose-built processors, embedded was first
swamped by commoditized x86 computing platforms and is now
on the cusp of inundation by another consumer platform wave.
Production volumes for applied (a.k.a. embedded) computing
are dominated today by long-lifecycle motherboards in kiosks,
digital signage, point of sale terminals and the like. Simple adaptations of processor vendor reference designs fueled the Asian
Invasion of embedded SBCs and modules at prices even below
the bill of materials costs of small form factor (SFF) embedded
computers from North American and European manufacturers. Hundreds of medical, industrial, transportation, energy and
military system OEMs benefited from pervasive standards-based
building blocks for their line-powered fixed always-on devices.
Though the bulk of system implementation resources shifted toward software, validation and regulatory tasks, there are still a
few unsung hardware heroes to give each system its unique application I/O capabilities.
Tablets and smartphones now threaten to claim a significant
number of embedded system designs. We’ve become spoiled by
the connectivity and convenience of these consumer and enterprise gadgets. Although still not suitable for applications requiring large-scale displays and high-end graphics, it’s hard to deny
the economies of scale and overall utility of these lightweight
ultra-portable gadgets for embedded applications. These rich
open platforms are rapidly replacing purpose-built ruggedized
user interface terminals for a variety of instruments and devices
in all market segments. In other words, “there’s an app for that!”
How can the SFF supplier community capitalize on this
trend? Clearly there is very little point in creating the design
equivalent of a phone or tablet. Popular x86 and Intel Architecture processors are already available in every conceivable shape
and size. So why not repurpose the same cookie cutter reference
design recipe for ARM modules? There is already some momentum in this direction, although the breakout of hype versus substance is yet to be seen. Besides low power, if the salient point
10
of using ARM is low cost, then why pay the overhead for COMs
with baseboards when a college student can implement the reference design on a single board in a few weeks? The real challenge and cost lies with the software. It’s strange that x86 board
vendors still need to learn what the RISC and microcontroller
custom design community has known for decades.
Perhaps the SFF community is better off enhancing the connectivity of its x86 boards to these ARM-based consumer devices.
Porting, testing and certifying operating systems and middleware
stacks creates value in saving the system OEM time and money.
Partnering or even merging with embedded software companies
would lead to a differentiated value proposition. Multicore processors can run the GUI and OS on one core and real-time operating
system (RTOS) on the other. System OEMs prefer suppliers who
stand behind their offering rather than pointing the finger at a third
party. Maybe the industry is ready for another wave of consolidation, this time involving hardware and software together rather
than only one or the other. Is the term “solution” overused yet?
Before settling into a funk, keep in mind that there is still the
core block of application-specific I/O functionality of the device
to develop or upgrade. Just a few examples include controlling a
laser or reading sensors, or capturing and compressing images,
or taking a measurement, or monitoring a production process or
other event-driven or real-time behavior. And for those design
elements, system OEMs still need to rely upon the gray-haired
or no-haired engineers among us. Those who remember that
embedded development environments once consisted of scopes,
logic analyzers, microprocessor emulators (with real bond-out
silicon), assemblers and debuggers with disassemblers and instruction trace capability. Code had to be written efficiently to
fit into kilobytes of RAM and mask ROM or EPROM. Fortunately, a few old-school designers and firmware engineers are
still around who know how to use these tools when exception
handlers or race conditions need to be debugged.
Of course, the job now is to hide the secret sauce behind
pretty APIs for the young buck software engineers. Who knows,
your medical surgery apparatus or weather balloon may have its
own following on Facebook or Twitter soon. Every silver lining
has a touch of gray.
YOUR OFFICE. OUR ROUTER.
Utilizing the same Cisco IOS software that IT staffs
are already trained on, our rugged, SWaP optimized
routers are battlefield-ready. The X-ES XPedite5205
PMC router module and the SFFR small form factor
router platform, working in conjunction with UHF, VHF, Wi-Fi and other
radio platforms create mobile ad hoc networks to provide highly secure
data, voice and video communications to stationary and mobile network
nodes across both wired and wireless links. Call us today to learn more.
Connected everywhere. That’s Extreme.
Extreme Engineering Solutions
608.833.1155 www.xes-inc.com
editor’s report
Realizing the Potential of Multicore Processors
Retooling Applications to
Ride on Multiple Cores
Taking advantage of the real potentials for performance
offered by multicore processors is an ongoing quest.
A new tool can dissect programs into multiple threads
that can run across multiple cores—with significant
performance gains.
by Tom Williams, Editor-in-Chief
I
t goes without saying that today multicore processors are all the rage. As silicon
manufacturers attempted to obtain increased performance by simply flogging the
clock faster, they ran up against the very real
limits of power consumption and heat dissipation. But advances in process technology and
reduced geometries have enabled us to put
multiple cores on the same silicon die and—
hopefully—parallelize our software so that it
effectively runs faster. But how successful has
that really been? And have the very perceptive performance gains really been indicative
of the potential that is there if it could be fully
exploited. We don’t really know.
Back in 2008, Steve Jobs, never one to
mince words, was quoted by the New York
Times as saying, “The way the processor
industry is going, is to add more and more
cores, but nobody knows how to program
those things. I mean, two, yeah; four, not really; eight, forget it.” Of course, Steve, that is
not going to stop us from trying. So among the
questions are how far have we come in exploiting this potential and where can we still go?
Getting the Juice out of Multicore
First of all, the utilization of multicore
architectures is taking place on several levels.
Among these are the operating system and a
variety of hypervisors, which are used to virtualize the multicore environment so that it can
be used by more than one operating system.
The next level is just starting to open up and,
12
of course, offers the biggest potential, which
is the ability of individual applications to run
smoothly across multiple processor cores.
At the operating system level, for example, Windows 7 running on a quad-core Intel
Core processors does take advantage of the
multiple cores to the extent it can. It can launch
an application on one core, drivers on another,
and it can assign a whole host of background
processes, of which the user is mostly never
aware, across other cores. The user has no
knowledge of or control over where these processes all run, but the fact that the operating
system can do this does significantly increase
the performance of the overall system.
The world of hypervisors comes in
several forms, which are basically software
entities that manage virtual machines so
that multiple guest operating systems can
run on the same hardware platform. John
Blevins, director of product marketing for
LynuxWorks, offers a scheme for classifying hypervisors based on their functionality.
There are Type 2 hypervisors that are basically emulation programs that run under a
host OS. This would allow the running of
a guest operating system such as Windows
under a host OS like Linux. Type 1 hypervisors are computer emulation software tightly
integrated with an OS to form a single “selfhosted” environment that runs directly on
the hardware. The lines between Type 1 and
Type 2 hypervisors can become a bit blurred.
Therefore, Blevins offers a third des-
ignation he calls the Type Zero hypervisor.
This type of hypervisor is built to contain
only the absolute minimum amount of software components required to implement
virtualization. These hypervisors are the
most secure type of hypervisor as they are
“un-hosted” and are dramatically smaller
than Type 1 and 2 hypervisors. Since the
Type Zero hypervisor runs directly on the
hardware below the operating system, it can
assign a given OS to a designated core or
cores giving that OS complete protection
from code running on other cores (Figure
1). Thus an OS that is multicore-aware could
run on its assigned cores just as if it were
on its own hardware platform. Windows 7,
for example, would function in such a hypervisor environment exactly as it does on
a multicore PC—within its assigned space.
None of this, however, addresses the need to
spread the actual applications across multiple cores to improve their performance.
Parallelizing the Application
In approaching the goal of actually getting everyday application programs to take
advantage of multiple cores, the object will
naturally be to first determine which parts
of the code can be targeted to run on independent cores while keeping their communication and synchronization with each other
and with the application as a whole. Multithreaded code offers itself to analysis and
parallelization, but the question of how to approach the task is nontrivial. Steve Jobs was
almost certainly correct in believing that nobody understands how to sit down at a computer and start writing multicore code from
scratch. You’ve got to establish the structure
of the program and then decide how to approach parallelization through analysis.
This is the approach of the Pareon tool
recently announced by Vector Fabrics. A
normal application that has not been parallelized will only run on one core. The
approach is to take the complete—or fairly
complete— written and debugged source
code and to understand it well enough to
pick out those elements that will actually
provide enhanced performance if they are
parallelized. That normally means that the
performance increase should significantly
exceed the amount of overhead involved in
parallelizing and such things as delays for
communication and synchronization.
editor’s report
PARTITION 0
PARTITION 1
PARTITION 2
PARTITION n
LynxOS-SE
RTOS
WINDOWS 7
WINDOWS XP
SECURE DEVICE
SERVER
DIRECT
DEVICE
ASSIGNMENT
VIRTUALIZED SHARED DEVICES
LynxSECURE
ETHERNET
SEPARATION KERNEL AND EMBEDDED HYPERVISOR
GRAPHICS
USB
SATA
PHYSICAL
DEVICE
ASSIGNMENT
ETHERNET 2
PHYSICAL DEVICES
Figure 1
The LynxSecure separation kernel/hypervisor from LynuxWorks is being called a Type Zero hypervisor because it runs directly on the hardware below the operating systems. It can assign a given OS to a core or cores and ensure that it runs exclusively on its assigned hardware.
This is fundamentally different from
what a hypervisor does. A hypervisor does
not touch the code but provides virtual
contexts in which it can run. Currently,
Pareon supports the x86 architecture and
the ARM Cortex A-9, mostly in a Linux
environment. It can work with C and C++
programs including programs that use binary libraries or access databases. It can
handle programs that include non-C/C++
codes such as assembler, but can only parallelize those sections written in C/C++.
The process starts by building the program and running it on a model of the target
hardware architecture. This allows insight
into deep details of behavior including such
things as bottlenecks, cache hits and misses,
bus bandwidth and traffic, memory access
times and more. After running this executable on the model, which is intended to give
deep understanding of how the processor
architecture and the code interact, Pareon
performs dynamic analysis to gain information that will support deciding how best to go
about an optimal parallelization. These are
the initial steps of a three-stage process consisting of what Vector Fabrics describes as,
“insight, investigation and implementation.”
Having done the basic analysis,
Pareon moves the user to the investigation
stage, which actually involves more deliberate analysis by the user of the data gathered during the insight phase. For example,
coverage analysis is needed to ensure that
Figure 2
This is a relatively complex loop with a number of dependencies, including
communication, but would gain significantly from parallelization. In the bar
above, the red shows the overhead that would be introduced compared to the
execution time. The speedup is indicated on the left.
the execution profile of the application is
as complete as possible so as not to miss
any parts that would gain significantly
from parallelization. In addition, a profiling analysis keeps track of where the ac-
tual compute cycles go. This can reveal
hotspots and provide a basis to explore and
find the best parallelization strategy.
For example, the user can select a given
loop and investigate the parallelization options.
13
editor’s report
FREE
DATA
ACQUISITION
HANDBOOK
Figure 3
While this loop could be parallelized, the view shows that the threads mostly wait on
each other and therefore would offer very little performance advantage if parallelized.
Learn about
DAQ,
A/D Converters,
Signal Conditioning,
Measurement
Techniques,
and more...
Download
Your Copy Today!
mccdaq.com
Contact us
1.80 0.234.4232
©2012 Measurement Computing Corporation
10 Commerce Way, Norton, MA 02766
[email protected]
14
Untitled-4 1
7/10/12 11:11 AM
The tool will alert the user to possible dependencies that would restrict partitioning. Otherwise, it presents options for loop parallelization from which the user can choose, such as
the number of threads. From this the tool can
show the overall performance increase, taking into account overhead involved with such
things as spawning and synchronizing tasks or
the effects of possible memory contention and
data communication (Figure 2).
Of course, not all loops are good candidates for parallelization. If threads have
to wait on each other so that one thread
must complete before the next can start,
parallelization is not going to be a great
benefit. So a good deal of the developer’s
effort is devoted to understanding the
code and making informed choices about
what to parallelize and what not to (Figure 3). To help with such decisions there is
a performance prediction feature that immediately shows the impact of parallelization on program performance.
Once the developer has gone through
and examined the parallelization options,
made his or her choices and gotten a view
of the overall improvement, it is time to
refactor the code to implement the chosen improvements. Pareon keeps track of
where the developer has chosen to add parallelism, and once the strategy has been selected presents the user with detailed stepby-step instructions on how to implement
that parallelism. Thus, there may not be an
automated button, but every step needed to
properly modify the source code is given.
Once the code is refactored, it can be
recompiled using a normal C or C++ compiler. In fact, Pareon integrates with standard
development tools so that familiar debuggers, profilers, text editors and other tools are
available for subsequent tweaks and examination of the code. Such steps will probably
be needed if the code is intended for embedded applications where timing constraints,
interrupt latencies and other issues affecting
deterministic behavior are critical.
In fact, Vector Fabrics is not currently
targeting the tool at hard real-time applications, but more at the tablet, handheld mobile and smartphone arena. The temptation
to use it for embedded applications will no
doubt be strong and it seems clear that this
avenue will also be pursued by other players
if not by Vector. The push to truly take advantage of the possibilities offered by multicore architectures is increasing—as is the
number of cores per die—and we can expect
further innovation in this dynamic arena.
LynuxWorks
San Jose, CA.
(408) 979-3900.
[www.lynuxworks.com].
Vector Fabrics
Eindhoven, Netherlands.
+31 40 820-0960.
[www.vectorfabrics.com].
USB Modules &
Data Acquisition Showcase
Featuring the latest in USB Modules &
Data Acquisition technologies
USB-104-HUB
Multifunction DAQ-PACK
System Series
Rugged, industrialized, four-port USB hub
Extended temperature operation
(-40°C to +85°C)
Supports bus powered and selfpowered modes
Three power input connectors (power
jack, screw terminals, or 3.5” drive
Berg power connector)
USB/104 form-factor for OEM
embedded applications
OEM version (board only) features
PC/104 module size and mounting
compatibility
Includes micro-fit embedded USB
header connectors
ACCES I/O Products, Inc.
Phone: (858) 550-9559
Fax: (858) 550-7322
Up to 128-channels, 16-bit singleended or differential analog inputs
Multifunction DAQ with sustained
sampling speeds up to 500kHz
Flexible, software configured
functionality
Wide variety of input ranges,
unipolar and bipolar, per 8-channel
programmable
Extensive range of flexible signal
conditioning types
ACCES I/O Products, Inc.
E-mail: [email protected]
Web: www.accesio.com
Phone: (858) 550-9559
Fax: (858) 550-7322
Get Connected
with ADLINK USB2401/1901/1902/1903
DAQ Modules!
ADLINK Technology, Inc.
Phone: (408) 360-0200
Fax: (408) 360-0222
16-bit, 16-CH, 250kS/s (USB-190x)
24-bit, 4-CH, 2kS/s (USB-2401)
Built-in signal conditioning enabling
direct measurement of current signal,
bridge-based sensor, TC and RTD.
(USB-2401)
USB powered
Easy device connection with removable
screw-down terminal module
Free U-test utility, delivering full
function testing with no programming
required
Web: www.adlinktech.com
Data Acquisition Handbook
The Data Acquisition Handbook
includes practical DAQ applications as
well as theoretical measurement issues.
Each chapter details basic DAQ and
signal conditioning principles including
sensor operation and the need for careful
system interconnections. Download your
FREE copy today!
Measurement Computing
Phone: (508) 946-5100
Fax: (508) 946-9500
USB Wi-Fi Modules
802.11b/g/n Compliant
Radicom Research, Inc.
Phone: (408) 383-9006
Fax: (408) 383-9007
USB 2.0 hot swappable interface
Compatible with USB1.1 and USB2.0
host controllers
Up to 300Mbps receive and 150Mbps
transmit rate using 40MHz bandwidth
Up to 150Mbps receive and 75Mbps
transmit rate using 20MHz bandwidth
1 x 2 MIMO technology for
exceptional reception and throughput
2 U.FL TX/RX antenna ports
Wi-Fi security using WEP, WPA,
WPA2
Compact size: 1.0” x 1.0” x 0.25”
(Modules)
Web: www.radi.com
Web: www.accesio.com
Web: www.mccdaq.com
Best Scope Selection
Saelig has the widest, unique selection
of scopes to suit any budget and
application: low-end PC scope adapters
up to high-end 16-bit 12GHz units.
Standalone, remarkable economical LCD
scopes, 7-in-1 toolbox essentials, up
to bargain-priced 4-channel benchtop
work horses with FFT and auto-test
capabilities. See our website for a great
selection of spectrum analyzers, RF
signal sources, and waveform generators
not available anywhere else! The best
selection and lowest prices are always
found at www.saelig.com
Saelig Company, Inc.
Phone: (585) 385-1750
Fax: (585) 385-1768
Web: www.saelig.com
ploration
your goal
k directly
age, the
source.
ology,
d products
Technology in
context
The Expanding Roles of Non-Volatile Memory
The Expanding Role of Non-Volatile
Memory in High-Performance
Embedded Architecture
The answer to the big question, “What will be the next non-volatile memory
to replace NAND flash?” is currently unclear, but due to the limitations of
NAND flash (endurance, reliability, speed) when compared to DRAM, it is
likely a new non-volatile memory technology will evolve.
by Adrian Proctor, Viking Technology
O
ver the past 20+ years, there have inate the computing industry today are,
been numerous memory tech- DRAM and NAND flash, but it should
nologies brought to market with be noted that both of these technolovarying degrees of commercial success. gies have their pros and cons (Table 1).
Among these are static RAM (SRAM), DRAM delivers the highest performance
nies providing
solutions
nowRAM, NOR flash, Eprom,
pseudo
static
(latency / speed), with practically infinite
ion into products,
technologies
and companies.
your goal
is to research
the latest yet it is volatile and has much
EEprom, DRAM
and Whether
NAND
flash.
endurance,
ation Engineer, or jump to a company's technical page, the goal of Get Connected is to put you
Generally speaking, these “memory” lower capacity points than other memoyou require for whatever type of technology,
ries such as NAND flash. On the other
and productstechnologies
you are searchingcan
for. be split into two categories, volatile and non-volatile. Vola- hand, NAND flash scales to high capactile memory will not retain data when ity, is non-volatile and relatively cheap
power is turned off; conversely, non-vol- ($//Gbit), but it is significantly slower
atile memory will. The two dominating than DRAM. Additionally, endurance
memory technologies in the industry and data retention are getting worse as
today are DRAM (volatile) and NAND process geometries continue to shrink,
flash (non-volatile). Figure 1 summa- meaning that for write-intensive enterrizes memory technologies as emerging, prise applications, NAND flash, in the
long term, may not be an optimal memniche and those in mass production.
The memory technologies that dom- ory technology.
There is much discussion in the
industry as to what new universal
Get Connected
memory technology or technologies
with companies mentioned in this article.
will materialize as real contenders to
www.rtcmagazine.com/getconnected
End of Article
16
Get Connected with companies mentioned in this article.
displace either or both NAND flash &
DRAM. Some of these newer emerging technologies include: Magnetic
RAM (MRAM), Ferroelectric RAM
(FRAM), Phase Change Memory
(PCM), Spin-Transfer Torque RAM
(STT-RAM) and Resistive RAM (ReRAM) or memristor. See sidebar
“Emerging Memory Technologies
Overview,” p. 22.
FRAM, MRAM and PCM are currently in commercial production, but
still, relative to DRAM and NAND
flash, remain limited to niche applications. There is a view that MRAM,
STT-RAM and ReRAM are the most
promising emerging technologies, but
they are still many years away from
competing for industry adoption. Any
new technology must be able to deliver
most, if not all of the following attributes in order to drive industry adoption on a mass scale: scalability of
´1RZLQWHOOLJHQWFRQQHFWHGGHYLFHV
FDQEHQHÀWPRUHWKDQHYHUIURP
,QWHO$WRPSURFHVVRUVµ
3UDQDY0HKWD
6HQLRU3ULQFLSDO(QJLQHHUDQG&72
,QWHO(PEHGGHGDQG&RPPXQLFDWLRQV*URXS
2QHVPDOO6R&
1RZHPEHGGHGZLWKELJJHULGHDV
'RPRUHZLWKWKH1HZ,QWHO$WRPSURFHVVRU([[VHULHV
EDVHGSODWIRUP1RZZLWKHQKDQFHG+'YLGHRFDSWXUHDQG
GLVSOD\FDSDELOLW\DQGVXSSRUWIRUPRUH26V
'XDO+''LVSOD\
+DUGZDUHDFFHOHUDWHGYLGHR
HQFRGLQJIRUGULYLQJKLJK
GHÀQLWLRQGXDOSGLVSOD\V
6$7$
([SDQGHGFRPSDWLELOLW\
ZLWK6HULDO$7$GHYLFHV
26
0LFURVRIW:LQGRZV(PEHGGHG
6WDQGDUG:LQGRZV(PEHGGHG
&RPSDFWDQG$QGURLG
:HWRRNRXUÀUVW,QWHO$WRPEDVHGSURFHVVRUDQGPDGHLWHYHQPRUHFDSDEOH$QGVWLOOUHWDLQHGWKH
ÁH[LELOLW\RIDVWDQGDUG3&,([SUHVVLQWHUFRQQHFWWKDWOHWV\RXSLFNDQGFKRRVH,2KXEVSURSULHWDU\
$6,&VRUGLVFUHWHGHYLFHV$VZHOODVVXSSRUWIRUHPEHGGHGRSHUDWLQJV\VWHPVLQFOXGLQJ$QGURLG
0HH*R)HGRUDDQG0LFURVRIW:LQGRZV;3(PEHGGHG6WDQGDUG(PEHGGHG&RPSDFW
DQG:LQGRZV(PEHGGHG6WDQGDUG
$OOLQDUXJJHGL]HGVPDOOIRRWSULQWSURFHVVRUZLWKDQLQGXVWULDOWHPSHUDWXUHUDQJH&WR&
DQGORZGHIHFWVSHUPLOOLRQ'30
7KLQNELJ*HWVWDUWHG
/($51025($7!HGFLQWHOFRPJRDWRPH[[

,QWHO&RUSRUDWLRQ,QWHOWKH,QWHOORJRDQGWKH,QWHO(PEHGGHGORJRDUHWUDGHPDUNVRI,QWHO&RUSRUDWLRQLQWKH86DQGRURWKHUFRXQWULHV6RIWZDUHDQGZRUNORDGVXVHGLQSHUIRUPDQFH
WHVWVPD\KDYHEHHQRSWLPL]HGIRUSHUIRUPDQFHRQO\RQ,QWHOPLFURSURFHVVRUV3HUIRUPDQFHWHVWVVXFKDV6<6PDUNDQG0RELOH0DUNDUHPHDVXUHGXVLQJVSHFLÀFFRPSXWHUV\VWHPVFRPSRQHQWV
VRIWZDUHRSHUDWLRQVDQGIXQFWLRQV$Q\FKDQJHWRDQ\RIWKRVHIDFWRUVPD\FDXVHWKHUHVXOWVWRYDU\<RXVKRXOGFRQVXOWRWKHULQIRUPDWLRQDQGSHUIRUPDQFHWHVWVWRDVVLVW\RXLQIXOO\HYDOXDWLQJ
\RXUFRQWHPSODWHGSXUFKDVHVLQFOXGLQJWKHSHUIRUPDQFHRIWKDWSURGXFWZKHQFRPELQHGZLWKRWKHUSURGXFWV2WKHUQDPHVDQGEUDQGVPD\EHFODLPHGDVWKHSURSHUW\RIRWKHUV
Sharpen your engineering skills
with Intel® at RTECC
Real-Time & Embedded
Computing Conference
August 21, 2012
Irvine, CA
Morning & Afternoon Sessions
Plus Hands-On Lab
Seating Is Limited
Intel® Boot Loader Development Kit (BLDK) for
Embedded Systems
Complete Agenda
Available Online
Start with an overview of BLDK and complete
your training with a Hands-on Lab
In the hands-on lab you will learn how to:
• Create a Boot Loader Development Kit (BLDK)
Project
• Build a Firmware Image Using Windows
Hosted Tools
• Boot an E6XX Systems to UEFI Shell &
Explore the Various Options
• Update E6XX Firmware from UEFI Shell
Register today at www.rtecc.com
See what’s on the schedule at
www.rtecc.com
August 21, 2012
Irvine, CA
Attendees who complete the class will be entered in a drawing
for an Intel® Atom™ Processor E6xx System (a $300 value)
?^acfT[[
LIGHTNING FAST BOOT
take advantage of the new Intel® BLDK
Intel® Boot Loader Technology
The Intel® Boot Loader Development Kit (Intel® BLDK) is a software toolkit that allows creation of
customized and optimized initialization firmware solutions for embedded Intel® processor-based platforms.
The Intel BLDK enables rapid development of firmware for fixed-function embedded designs—those requiring
basic initialization and functionality rather than the full capabilities delivered with a traditional BIOS.
Support Intel ® BLDK technology
Intel ® Atom™ E6X0T processor
Dual display: VGA and 18/24bit LVDS
1 x SD Socket, 1 x Mini PCI-E and
1 x PCI-E expansion slot
Support up to 2GB DDR2 SDRAM
Small Form Factor
120(L) x 120(W) mm

NANO-6040 Nano-ITX Board
Support Intel ® BLDK technology
Intel ® Atom™ E6X0T processor
Single channel display with LVDS
4 x PCI Express x1 expansion slot
CAN Bus interface supported
Support up to 2GB DDR2 SDRAM
Small Form Factor
70(L) x 70(W) mm
PQ7-M105IT Qseven Module
Portwell Industrial Embedded Solutions
Wide selection of Mini-ITX, Nano-ITX, 3.5” ECX, COM Express
and Qseven modules
The latest Intel® Core™ i5/i7 and single-/
dual-core Atom™ CPUs
Custom carrier board design and manufacturing
Quick time-to-market solutions
Extended temp and ECC memory support
Small Form Factor
ISO 9001:2008
ISO 13485:2003
technology in context
Emerging Memory
Technologies Overview
“Memory”
MRAM: Magnetic RAM
MRAM is a non-volatile memory. Unlike DRAM,
the data is not stored in electric charge flows,
but by magnetic storage elements. The storage elements are formed by two ferromagnetic
plates, each of which can hold a magnetic field,
separated by a thin insulating layer. One of the
two plates is a permanent magnet set to a particular polarity; the other’s field can be changed to
match that of an external field to store memory.
Vendors: Everspin (Freescale spin-off), Crocus
Technology
STT-RAM:
RAM
Spin-Transfer
Torque
STT-RAM is an MRAM, which is non-volatile, but
with better scalability over traditional magnetic
RAM. STT is an effect in which the orientation of
a magnetic layer in a magnetic tunnel junction or
spin valve can be modified using a spin-polarized
current. Spin-transfer torque technology has the
potential to make possible MRAM devices combining low current requirements and reduced
cost. However, the amount of current needed to
reorient the magnetization is at present too high
for most commercial applications.
Vendors: Samsung, SK-Hynix, Renesas,
Toshiba, Everspin, Crocus Technology
PCM: Phase Change Memory
PCM is a non-volatile random access memory.
It utilizes the unique behavior of chalcogenide—
a material that has been used to manufacture
CDs—whereby the heat produced by the passage of an electric current switches this material
between two states. The different states have
different electrical resistance, which can be used
to store data. It is expected PCM will have better
scalability than other emerging technologies.
Vendors: Micron, Samsung
ReRAM: Resistive RAM
ReRAM is a non-volatile memory that is similar to PCM. The technology concept is that a
dialectric, which is normally insulating, can be
made to conduct through a filament or conduction path formed after application of a sufficiently high voltage. Arguably, this is a memristor technology and should be considered
as potentially a strong candidate to challenge
NAND flash.
Vendors: SK-Hynix, HP, NEC, Panasonic,
Samsung
22
Volatile
Memory
Non-Volatile
Memory
Mass
Production
Emerging
Technology
Mass
Production
Niche
Emerging
Technology
DRAM
Floating
Body
NAND
MRAM
ReRAM
SRAM
NOR
FRAM
STT-RAM
Psuedo
SRAM
EPROM
PCM
EEPROM
Racetrack
MRAM
Figure 1
Memories categorized as mass production, niche application or emerging
technology.
the technology, speed of the device,
power consumption better than existing
memories, endurance, densities better
than existing technologies, and finally
cost—if the emerging technology can
only manage one or two of these attributes, then, at best, it is likely to be resigned to niche applications.
So the answer to the question, “What
will be the next non-volatile memory to
replace NAND flash?” is almost certainly that a new non-volatile memory
technology will evolve. However, it probably will not replace the current mainstream memories for at least the next 5 to
7 years. However, non-volatile DIMMs,
such as the ArxCis-NV from Viking
Technology, can enable increased application performance and far improved
power failure / system recovery, when
compared to current implementations
(Figure 2).
What Is a Non-Volatile DIMM?
A non-volatile DIMM is a module
that can be integrated into the main memory of high-performance embedded compute platforms, such as AdvancedTCA
Blades; perform workloads at DRAM
speeds; yet be persistent and provide data
retention in the event of a power failure or
system crash.
A non-volatile DIMM is a memory
subsystem that combines the speed and
endurance of DRAM together with the
non-volatile data retention properties of
NAND flash. This marriage of DRAM
and NAND technology delivers a highspeed and low-latency “non-volatile /
persistent” memory module. Designed
from the ground up to support unlimited read/write activity, it performs at
fast DDR3 speeds and can sustain itself
in the face of host power failure or a system crash.
Figure 2
The Viking Arx-Cis-NV fits into a socket in main memory but has power backup
via a supercapacitor, making it a non-volatile DRAM.
Server performance gap between technologies
Performance
What makes these modules different from standard DRAM modules is that
in the event of a power failure or system
crash, the data in the NV-DIMM is securely preserved and available almost
immediately upon power being restored
to the host system, much as in the case
of suspend/resume. If performance is
critical to business success and if minimizing downtime is an important issue,
then NV-DIMMs will be extremely valuable wherever the application bottleneck
is storage and I/O, and where downtime
costs money.
With the recent surge in use of
SSDs in high-performance compute
environments, and those architectures also utilizing caching software
(auto-tiering), many applications have
enjoyed significant performance improvements. Therefore the SSD and
software bundle has significantly improved the memory/storage gap, which
has helped alleviate I/O bottlenecks
(Figure 3).However, the fact remains
that NAND flash SSDs are best suited
for read applications, not write-intensive ones. Thus, these intelligent caching software solutions, when paired
with SSDs, will utilize other system
resources to ensure performance & reliability, i.e., CPU, DRAM and HDDs.
Indeed most SSD caching software
will utilize the host system’s standard
DRAM for intensive write activity to
preserve the SSD and also keep the write
cache in DRAM. That means this critical
data is susceptible to loss in the event of
a system crash or power failure. A common work-around to protect the data is
to “checkpoint” the write buffer out into
slower block-based storage, namely disk,
at regular intervals. But “checkpointing”
naturally has a negative impact on the I/O
performance.
With NV-DIMMs integrated into the
host system, data-centric and write-intensive applications will enjoy a significant
increase in performance. In addition to
the benefits of increased application performance, recovery is another area of significant benefit.
Should the system experience a
power outage, one of two scenarios
Time
Server CPU
Speeds
Caching Software
w/SSDs does
better job of
bridging I/O gap
SSD Tier 0
SSDs implemented
as a higher tier of
storage improves
I/O performance
HDD
1970
1980
1990
2000
2012
Figure 3
will occur. Either the power failure will
cause a catastrophic loss of the “inmemory state,” or the backup power supplies will enable the appliances to transfer this data held in main memory out
to disk. The entire state—which could
be hundreds of gigabytes of DRAM—
must be saved to a storage back end.
Both “saving” and “recovering/reconstructing” this amount of data, across
multiple servers, will be extremely slow
and place a very heavy load on the storage infrastructure, resulting in severe
I/O bottlenecks.
When
utilizing
non-volatile
DIMMs, in the event of power loss or
if the system crashes unexpectedly,
the NV-DIMMs allow the system to
recover their “in-memory state” almost instantaneously without putting
any load on the storage back end—in
a sense, making the failure appear as a
23
suspend/resume event. This means that
business-critical applications such as
online transaction processing (OLTP)
can be up and running again in a matter
of minutes rather than hours and without the need for uninterruptable power
supply (UPS) intervention.
Pro’s & Con’s of DRAM & NAND Flash
Technology
DRAM
NAND
Speed
×
Power
×
×
Capacity (GB)
Endurance
×
Data Retention (Non-Volatile)
×
Cost/GB
×
Reliability
×
Technical Improvements with
Geometry Shrinks
×
TABLE 1
24
Untitled-7 1
The Role of the Supercapacitor
In order for a non-volatile DIMM
to perform its task, a small energy
source is required to ensure 100% data
security on system failure. Supercapacitors are an appropriate technology
for use in this environment, primarily
because they provide a superior solution when compared to batteries. Any
technology, when new and relatively
unknown, will encounter questions
about long-term reliability and capabilities. There are a number of reasons
why supercapacitors are suitable for
use in this type of application.
Supercapacitors are highly efficient
components with high current capability. Their efficiency—defined as the to-
6/7/12 3:34 PM
tal charge removed divided by the total
charge added to replenish the charge
removed—is greater than 99%, even at
very high currents, meaning that little
charge is lost when charging and discharging the supercapacitor. Since supercapacitors are designed with a very
low equivalent series resistance (ESR),
they are able to deliver and absorb very
high current. The inherent characteristics of the supercapacitor allow it to
be charged and discharged at the same
rates, something no battery can tolerate.
In battery-based systems, you can only
charge as fast as the battery will accept
the charge.
Since supercapacitors operate without relying on chemical reactions, they
can operate over a wide range of temperatures. On the high side, they can operate
up to 65°C, and withstand storage up to
85°C, without risk of thermal runaway.
On the low side, they can deliver power as
cold as -40°C.
Determining battery state of charge
(SOC) and state of health (SOH) is a
significant consideration for robust battery systems, requiring sophisticated
data acquisition, complex algorithms
and long-term data integration. In comparison, it is very simple to determine
the SOC and SOH of supercapacitors.
At the same time, the energy storage mechanism of a supercapacitor is
capable of hundreds of thousands of
complete cycles with minimal change
in performance. They can be cycled infrequently, where they may only be discharged a few times a year, or they may
be cycled very frequently.
Life cycle and maintenance are
also important factors. The energy
storage mechanism of a supercapacitor is a very stable process. It is capable of many years of continuous duty
with minimal change in performance.
In most cases, supercapacitors are installed for the life of the system. In addition, supercapacitors cannot be over
charged/discharged, and can be held at
any voltage at or below their rating. If
kept within their wide operating ranges
of voltage and temperature, there is no
recommended maintenance.
delivers value that far surpasses a simple
DRAM, DIMM and SSD architecture; it
is greater than the sum of the two technologies used.
The architecture of NV-DIMMs
provides a full interconnect on-module
that will independently allow transfer of data between the DRAM and the
flash without contention for other I/O
or CPU resources. Ultimately, applications can rely on the high-speed memory
(DRAM) to be “persistent” and need not
slow down to “checkpoint” or consume
other system resources. The NV-DIMM
Viking Technology
Foothill Ranch, CA.
(949) 643-7255.
[www.vikingtechnology.com].
TRRUST-Stor
™
A New Level of SSD Security
8JUIFODSZQUJPOVOQBSBMMFMFE
SVHHFEJ[BUJPOBOECMB[JOHGBTU
FSBTF.JDSPTFNJT5336454UPS™
JTUIFmSTUBOEPOMZTPMJETUBUFESJWF
EFTJHOFEGSPNUIFHSPVOEVQGPS
IJHIFOEJOEVTUSJBMBFSPTQBDF
EFGFOTFBOEJOUFMMJHFODFBQQMJDBUJPOT
t PS(#EFOTJUJFT
t )BSEXBSFJNQMFNFOUFE"&4FODSZQUJPO
t 5BNQFSSFTJTUBOUGFBUVSFTBWBJMBCMF
t &0-NBOBHFNFOU
t 5336451VSHF™SFOEFSTEBUBOPOSFDPWFSBCMFJONJMMJTFDPOET
t 'VMMESJWFFSBTFJOMFTTUIBOTFDPOET
t )BSEXBSFCBTFEBVUIFOUJDBUJPO
t -PXQPXFSPQFSBUJPO
t .5#'HSFBUFSUIBOIPVST
t .JMJUBSZHPWFSONFOUBHFODZTBOJUJ[BUJPOQSPUPDPMT
t #VJMU*O4FMG5FTU#*45
Email: [email protected]
website: www.microsemi.com
Power Matters.
Untitled-2 1
25
6/6/12 3:52 PM
RTC MAGAZINE JULY
2012
technology in
systems
OpenCL Programming: Parallel
Processing Made Faster and Easier
than Ever
Newer processors may not only have multiple cores of the same
architecture, they may also integrate heterogeneous computing elements.
Programming such devices with a single code base has just gotten easier.
by Todd Roberts, AMD
P
arallel processing isn’t really new.
It has been around in one form or
another since the early days of computing. As traditional CPUs have become
multicore parallel processors, with many
cores in a socket, it has become more important for developers to embrace parallel processing architectures as a means
to realize significant system performance
improvements. This move toward parallel
processing has been complicated by the
diversity and heterogeneity of the various
parallel architectures that are now available. A heterogeneous system is made up
of different processors, each with specialized capabilities. Over the last several
years GPUs have been targeted as yet another source of computing power in the
system. GPUs, which have always been
very parallel, counting hundreds of parallel execution units on a single die, have
now become increasingly programmable,
to the point that it is now often useful to
think of GPUs as many-core processors
instead of special purpose accelerators.
26
All of this diversity has been reflected
in a wide array of tools and programming
models required for programming these architectures. This has created a dilemma for
developers. In order to write high-performance code they have had to write their code
specifically for a particular architecture and
give up the flexibility of being able to run on
different platforms. In order for programs to
take advantage of increases in parallel processing power, however, they must be written in a scalable fashion. Developers need
the ability to write code that can be run on
a wide range of systems without having to
rewrite everything for each system.
OpenCL for Unified, Portable
Source Code
OpenCL, the first open and royalty-free
programming standard for general-purpose
parallel computations on heterogeneous systems, is quickly growing in popularity as a
means for developers to preserve their expensive source code investments and easily
target multicore CPUs and GPUs.
OpenCL is maintained by the Khronos
Group, a not-for-profit industry consortium
that creates open standards for the authoring and acceleration of parallel computing,
graphics, dynamic media, computer vision
and sensor processing on a wide variety
of platforms and devices. Developed in
an open standards committee with representatives from major industry vendors,
OpenCL affords users a cross-vendor, nonproprietary solution for accelerating their
applications across mainstream processing
platforms, and provides the means to tackle
major development challenges, such as
maximizing parallel compute utilization,
efficiently handling data movement and
minimizing dependencies across cores.
Ultimately, OpenCL enables developers to focus on applications, not just chip
architectures, via a single, portable source
code base. When using OpenCL, developers
can use a unified tool chain and language to
target all of the parallel processors currently
in use. This is done by presenting the developer with an abstract platform model that
tech in systems
conceptualizes all of these architectures in
a similar way, as well as an execution model
supporting data and task parallelism across
heterogeneous architectures.
Write A
Write B
Key Concepts and Workflows
OpenCL has a flexible execution
model that incorporates both task and data
parallelism (see sidebar “Task Parallelism
vs. Data Parallelism”). Tasks themselves
are comprised of data-parallel kernels,
which apply a single function over a range
of data elements in parallel. Data movements between the host and compute devices, as well as OpenCL tasks, are coordinated via command queues. Where the
concept of a kernel usually refers to the
fundamental level of an operating system,
here the term identifies a piece of code that
executes on a given processing element.
An OpenCL command queue is created by the developer through an API call,
and associated with a specific compute
device. To execute a kernel, the kernel is
pushed onto a particular command queue.
Enqueueing a kernel can be done asynchronously, so that the host program may
enqueue many different kernels without
waiting for any of them to complete. When
enqueueing a kernel, the developer optionally specifies a list of events that must occur
before the kernel executes. If a developer
wishes to target multiple OpenCL compute devices simultaneously, the developer
would create multiple command queues.
Command queues provide a general way
of specifying relationships between tasks,
ensuring that tasks are executed in an order
that satisfies the natural dependences in the
computation. The OpenCL runtime is free to
execute tasks in parallel if their dependencies
are satisfied, which provides a general-purpose task parallel execution model.
Write C
Kernel A
Kernel C
Kernel B
Read A
Kernel D
Read B
Figure 1
Task parallelism within a command queue.
Events are generated by kernel completion, as well as memory read, write and
copy commands. This allows the developer
to specify a dependence graph between
kernel executions and memory transfers in
a particular command queue or between
command queues themselves, which the
OpenCL runtime will traverse during execution. Figure 1 shows a task graph illustrating the power of this approach, where
arrows indicate dependencies between
tasks. For example, Kernel A will not execute until Write A and Write B have fin-
ished, and Kernel D will not execute until
Kernel B and Kernel C have finished.
The ability to construct arbitrary task
graphs is a powerful way of constructing taskparallel applications. The OpenCL runtime
has the freedom to execute the task graph in
parallel, as long as it respects the dependencies encoded in the task graph. Task graphs
are general enough to represent the kinds
of parallelism useful across the spectrum of
hardware architectures, from CPUs to GPUs.
Besides the task parallel constructs provided in OpenCL, which allow synchronizaRTC MAGAZINE JULY 2012
27
Tech In Systems
Sx
Task Parallelism vs.
Data Parallelism
OpenCL supports task-parallel and dataparallel programming models, each optimized
for different types of problems and processing
platforms.
Task parallelism is the simultaneous execution on multiple cores of many different
functions across the same or different datasets, and is ideally suited for multicore CPUs. In
this model, an instance of code is executed on
a device independent of other operations being
executed on other devices. Traditionally this has
been in the form of a thread of code running on
a CPU core. In OpenCL this can be a kernel executing on the CPU or GPU. Parallelism occurs
when multiple threads or kernels are executing
at the same time.
Data parallelism is the simultaneous execution on multiple cores of the same function
across the elements of a dataset, and is ideally
suited for GPUs. In the data-parallel programming model, a computation is defined in terms
of a sequence of instructions executed on multiple elements of a memory object. These elements are typically arranged in an index space,
which defines how the execution maps onto the
work items.
tion and communication between kernels,
OpenCL supports local barrier synchronizations within a work group. This mechanism
allows work items to coordinate and share
data in the local memory space using only
very lightweight and efficient barriers. Work
items in different work groups should never
try to synchronize or share data, since the
runtime provides no guarantee that all work
items are concurrently executing, and such
synchronization easily introduces deadlocks.
Developers are also free to construct
multiple command queues, either for parallelizing an application across multiple
compute devices, or for expressing more
parallelism via completely independent
streams of computation. OpenCL’s ability
to use both data and task parallelism simultaneously is a great benefit to parallel
application developers, regardless of their
intended hardware target.
Kernels
As mentioned, OpenCL kernels
provide data parallelism. The kernel ex-
28
work group (Wx , Wy)
Gx
sx = 0
sy = 0
work item
(WxSx + sx ,
WySy + sy)
sx = Sx - 1
sy = 0
work item
(WxSx = sx ,
WySy + sy)
Sy
Gy
sx = 0
sy = Sy - 1
work item
(WxSx = sx ,
WySy + sy)
sx = Sx - 1
sy = Sy - 1
work item
(WxSx + sx ,
WySy + sy)
Figure 2
Executing kernels - work groups and work items.
ecution model is based on a hierarchical abstraction of the computation being
performed. OpenCL kernels are executed
over an index space, which can be 1, 2 or 3
dimensional. In Figure 2, we see an example of a 2-dimensional index space, which
has Gx * Gy elements. For every element
of the kernel index space, a work item will
be executed. All work items execute the
same program, although their execution
may differ due to branching based on data
characteristics or the index assigned to
each work item.
The index space is regularly subdivided into work groups, which are tilings
of the entire index space. In Figure 2, we
see a work group of size Sx * Sy elements.
Each work item in the work group receives
a work group ID, labeled (wx, wy) in the
figure, as well as a local ID, labeled (sx,
sy) in the figure. Each work item also receives a global ID, which can be derived
from its work group and local IDs.
Work items in different work groups
may coordinate execution through the use
of atomic memory transactions, which are
an OpenCL extension supported by some
OpenCL runtimes. For example, work
items may append variable numbers of results to a shared queue in global memory.
However, it is good practice that work
items do not, generally, attempt to communicate directly because without careful
design, scalability and deadlock can be-
voidtrad_mul(int n,
const float *a,
const float *b,
float *c)
{
int i;
for (i=0; i<n; i++)
c[i] = a[i] * b[i]; }
Figure 3
Example of traditional loop
(scalar).
come difficult problems. The hierarchy of
synchronization and communication provided by OpenCL is a good fit for many of
today’s parallel architectures, while still
providing developers the ability to write
efficient code, even for parallel computations with non-trivial synchronization and
communication patterns.
The work items may only communicate and synchronize locally, within a
work group, via a barrier mechanism. This
provides scalability, traditionally the bane
of parallel programming. Because communication and synchronization at the finest granularity are restricted in scope, the
OpenCL runtime has great freedom in how
work items are scheduled and executed.
A Typical OpenCL Kernel
As already discussed, the core programming goal of OpenCL is to provide
tech in systems
programmers with a data-parallel execution model. In practical terms this means
that programmers can define a set of instructions that will be executed on a large
number of data items at the same time.
The most obvious example is to replace
loops with functions (kernels) executing
at each point in a problem domain.
Referring to Figures 3 and 4, let’s
say you wanted to process a 1024 x 1024
image (your global problem dimension).
You would initiate one kernel execution
per pixel (1024 x 1024 = 1,048,576 kernel executions).
Figure 3 shows sample scalar code
for processing an image. If you were writing very simple C code you would write a
simple for loop, and in this for loop you
would go from 1 to N and then perform
your computation.
An alternate way to do this would be
in a data parallel fashion (Figure 4), and
in this case you’re going to logically read
one element in parallel from all of a (*a),
multiply it from an element of b in parallel and write it to your output. You’ll notice that in Figure 4 there is no for loop—
you get an ID value, read a value from a,
multiply by a value from b and then write
the output.
As stated above, a properly written
OpenCL application will operate correctly on a wide range of systems. While
this is true, it should be noted that each
system and compute device available to
OpenCL may have different resources
and characteristics that allow and sometimes require some level of tuning to
achieve optimal performance. For example, OpenCL memory object types
and sizes can impact performance. In
most cases key parameters can be gathered from the OpenCL runtime to tune
the operation of the application. In addition, each vendor may choose to provide extensions that provide for more
options to tune your application. In most
cases these are parameters used with the
OpenCL API and should not require extensive rewrite of the algorithms.
Building an OpenCL Application
An OpenCL application is built by
first querying the runtime to determine
which platforms are present. There can
be any number of different OpenCL
A minimalist OpenCL program
#include <CL/cl.h>
#include <stdio.h>
#define NWITEMS 512
A simple memset kernel
const char *source =
“__kernel void memset( __global uint *dst )
“{
“ dst[get_global_id(0)] = get_global_id(0);
“}
\n”
\n”
\n”
\n”;
int main(int argc, char ** argv)
{
1. Get a platform.
cl_platform_id platform;
clGetPlatformIDs( 1, &platform, NULL );
2. Find a gpu device.
cl_device_id device;
clGetDeviceIDs( platform, CL_DEVICE_TYPE_GPU,
1,
&device
NULL);
3. Create a context and command queue on that device.
cl_context context = clCreateContext( NULL,
1,
&device,
NULL, NULL, NULL);
cl_command_queue queue = clCreateCommandQueue( context,
device,
0, NULL );
4. Perform runtime source compilation, and obtain kernel entry point.
cl_program program = clCreateProgramWithSource( context,
1,
&source,
NULL, NULL );
clBuildProgram( program, 1, &device, NULL, NULL, NULL );
cl_kernel kernel = clCreateKernel( program, “memset”, NULL );
5. Create a data buffer.
cl_mem buffer = clCreateBuffer( context,
CL_MEM_WRITE_ONLY,
NWITEMS * sizeof(cl_uint),
NULL, NULL );
6. Launch the kernel. Let OpenCL pick the local work size.
size_t global_work_size = NWITEMS;
clSetKernelArg(kernel, 0, sizeof(buffer), (void*) &buffer);
clEnqueueNDRangeKernel( queue,
kernel,
1,
NULL,
&global_work_size,
NULL, 0, NULL, NULL);
clFinish( queue );
7. Look at the results via synchronous buffer map.
cl_uint *ptr;
ptr = (cl_uint *) clEnqueueMapBuffer( queue,
buffer,
CL_TRUE,
CL_MAP_READ,
0,
NWITEMS * sizeof(cl_uint),
0, NULL, NULL, NULL );
int i;
for(i=0; i < NWITEMS; i++)
printf(“%d %d\n”, i, ptr[i]);
return 0;
}
CODE BLOCK 1
29
Tech In Systems
kernel void
dp_mul (global const float *a,
global const float *b,
global float *c)
{
int id = get_global_id (0);
c[id] = a[id] * b[id];
} // execute over “n” work-items
Figure 4
Data parallel OpenCL.
implementations installed on a single
system. The desired OpenCL platform
can be selected by matching the platform vendor string to the desired vendor
name, such as “Advanced Micro Devices, Inc.” The next step is to create a
context. An OpenCL context has associated with it a number of compute devices
(for example, CPU or GPU devices).
Within a context, OpenCL guarantees
a relaxed consistency between these devices. This means that memory objects,
such as buffers or images, are allocated
per context; but changes made by one
device are only guaranteed to be visible
by another device at well-defined synchronization points. For this, OpenCL
provides events with the ability to syn-
chronize on a given event to enforce the
correct order of execution.
Most OpenCL programs follow the
same pattern. Given a specific platform,
select a device or devices to create a context, allocate memory, create device-specific command queues, and perform data
transfers and computations. Generally,
the platform is the gateway to accessing
specific devices, given these devices and a
corresponding context. The application is
independent of the platform. Given a context, the application can:
• Create one or more command queues.
• Create programs to run on one or more
associated devices.
• Create kernels within those programs.
• Allocate memory buffers or images,
either on the host or on the device(s)—
Memory can be copied between the
host and device.
• Write data to the device.
• Submit the kernel (with appropriate
arguments) to the command queue for
execution.
• Read data back to the host from the
device.
ATCA, μTCA, VME AND
VPX SYSTEMS...FASTER.
Propel your project success with Schroff ® Systems and Subracks EXPRESS. We
provide VITA and PICMG compliant product solutions faster and at a competitive price.
Protect your application with standard or customized electro-mechanical and system
products – shipped in as few as two weeks and backed by our global
network and more than 60 years of engineering experience.
See our complete offering online.
RAPID DELIVERY
VITA and PICMG compliant solutions.
30
Untitled-6 1
W W W.SCHROFF.US
7/9/12 4:04 PM
tech in systems
The relationship between context(s),
device(s), buffer(s), program(s), kernel(s)
and command queue(s) is best seen by
looking at sample code.
Example Program – Simple
Buffer Write
Here is a simple programming example—a simple buffer write—with explanatory comments.
This code sample shows a minimalist OpenCL C program that sets a given
buffer to some value. It illustrates the basic programming steps with a minimum
amount of code. This sample contains
no error checks and the code is not generalized. Yet, many simple test programs
might look very similar. The entire code
for this sample is provided in Code Block
1.
1. The host program must select a
platform, which is an abstraction for a
given OpenCL implementation. Implementations by multiple vendors can coexist on a host, and the sample uses the first
one available.
2. A device ID for a GPU device is requested. A CPU device could be requested
by using CL_DEVICE_TYPE_CPU instead.
The device can be a physical device, such
as a given GPU, or an abstracted device,
such as the collection of all CPU cores on
the host.
3. On the selected device, an OpenCL
context is created. A context ties together
a device, memory buffers related to that
device, OpenCL programs and command
queues. Note that buffers related to a device can reside on either the host or the device. Many OpenCL programs have only
a single context, program and command
queue.
4. Before an OpenCL kernel can be
launched, its program source is compiled,
and a handle to the kernel is created.
5. A memory buffer is allocated on
the device.
6. The kernel is launched. While it is
necessary to specify the global work size,
OpenCL determines a good local work
size for this device. Since the kernel was
launched asynchronously, clFinish() is
used to wait for completion.
7. The data is mapped to the host for
examination. Calling clEnqueueMapBuffer ensures the visibility of the buffer
on the host, which in this case probably
includes a physical transfer. Alternatively,
we could use clEnqueueWriteBuffer(),
which requires a pre-allocated host-side
buffer.
OpenCL affords developers an elegant, non-proprietary programming
platform to accelerate parallel processing
performance for compute-intensive applications. With the ability to develop and
maintain a single source code base that
can be applied to CPUs, GPUs and APUs
with equal ease, developers can achieve
significant programming efficiency gains,
reduce development costs, and speed their
time-to-market.
Advanced Micro Devices
Sunnyvale, CA.
(408) 749-4000.
[www.amd.com].
We can help keep your VME
dinosaur running...even better.
Feed your beast with our embedded legacy VME upgrades
and keep your system from going extinct.
XV
6U ME-6
Pro 30
ces 0 i7
sor Int ®
Mo el
dul Cor ™
e
e
- Plug into any VME legacy system
(including 5V only chassis)
- Hardware byte swap
- 3 or 5 row VME connectors,
with or without P0
- Extensive I/O options
EMBEDDED
AN ACROMAG, INC. GROUP
90(93;F3&,30&&20([SUHVV
Untitled-3 1
www.xembedded.com
734-975-0577
31
7/10/12 11:07 AM
RTC MAGAZINE JULY
2012
technology in
systems
Developing Embedded Hybrid Code
Using OpenCL
Open Computing Language (OpenCL) is a specification and a programming
framework for managing heterogeneous computing cores such as CPUs
and graphics processing units (GPUs) to accelerate computationally
intensive algorithms.
by Mark Benson, Director of Software Strategy, Logic PD
I
32
NDRange size Gx
OpenCL NDRange (N=2)
Work-item
Work-item
(sx,sy) = (0,0)
(sx,sy) = (Sx-1,0)
Work-item
Work-item
(sx,sy) = (0, Sy-1)
(sx,sy) = (Sx-1,Sy-1)
Work-group size Sy
Work-group (Wx, Wy)
NDRange size Gy
n recent years, the mechanism by
which incremental computational performance has been achieved has shifted from clock speed to a proliferation of
processing cores. This shift, being driven
primarily by undesirable quantum effects
at higher signaling speeds and practical
limits on the rates we can dissipate heat,
has caused an acceleration of new software techniques. These techniques allow
us to not only leverage homogeneous multicore CPUs, but also graphics accelerators, digital signal processors (DSPs) and
field-programmable gate arrays (FPGAs)
as general-purpose computing blocks to
accelerate algorithms hungry for everhigher computational performance.
Proposed by Apple and maintained by
the Khronos Group, OpenCL was created
to provide a portable open programming
framework that enables software to take
advantage of both multicore CPUs and also
specialized processing cores, most notably
GPUs, for non-graphical processing purposes in a highly parallel way.
OpenCL is similar to OpenGL in
that it is a device-agnostic open standard
that anyone can adopt and use to create a
custom implementation. OpenCL was designed to work with OpenGL in that data
can be shared between frameworks—data
can be crunched with OpenCL and subsequently displayed using OpenGL.
Work-group size Sx
Figure 1
An OpenCL NDRange.
The OpenCL specification was developed by a working group formed in 2008,
chaired by Nvidia, and edited by Apple.
Since then, backward-compatible revisions
of the OpenCL specification have been released along with a set of conformance tests
that can be used to demonstrate compliance.
Conformant implementations of
OpenCL for a given processor are available primarily from the silicon vendor (Altera, AMD, ARM, Freescale, Imagination
Technologies, Intel, Nvidia, Texas Instruments, Xilinx, etc.). An OpenCL driver
from these vendors is required in order for
the OpenCL framework to run on top of it.
OpenCL is similar to Nvidia’s
CUDA, Brook from Stanford and Microsoft DirectCompute. In relation to these,
OpenCL has a reputation of being open,
portable, lower-level, closer to the hardware, and in some ways harder to use.
Think of OpenCL as a portable hardware
abstraction layer that supports parallel
programming on heterogeneous cores.
OpenCL also comes with a language
that is based on a subset of C99 with some
Medical Electronic Devices Are Changing the Way We Live.
Be Part of the Action!
BRINGING TOGETHER
ENGINEERS & DEVELOPERS
TO LEARN, ENGAGE
AND COLLABORATE ON
SOLUTIONS TO POWER THE
NEXT ERA OF HEALTHCARE.
Free Technical Conference
· Keynote Address
· Exhibition
· Seminars/Workshops
· New Product Showcase
· Networking
Upcoming Event:
August 23rd San Diego, CA
Colocated with the
REAL-TIME & EMBEDDED
COMPUTING CONFERENCE
Copyright © 2012 RTD Embedded Technologies, Inc. All rights reserved.
All trademarks or registered trademarks are the property of their respective companies.
PCI Express, PCI, and ISA Experts
RTD Offers In-House Design, Engineering, Manufacturing & Technical Support
4JOHMF#PBSE$PNQVUFST
t
t
t
t
t
*OUFM$PSF%VP1FOUJVN.$FMFSPO.
-PX1PXFS".%(FPEF-9
3VHHFE4VSGBDF.PVOU4PMEFSFE3".
0OCPBSE*OEVTUSJBM'MBTI%JTL
UP$0QFSBUJPO
%BUB$PMMFDUJPO.PEVMFT
t
t
t
t
t
t
4QFDJBMUZ.PEVMFT
t
t
t
t
t
t
1PSU&UIFSOFU4XJUDI
%FMUB4JHNB"OBMPH*0
)PU4XBQQBCMF3FNPWBCMF4"5"
6TFS$POGJHVSBCMF.JOJ1$*F
%JHJUBM4JHOBM1SPDFTTPST
t
"VUP$BMJCSBUJOH"OBMPH*0
"EWBODFE%JHJUBM*0
4JNVMUBOFPVT4BNQMJOH
)JHI4QFFE.D#41
1VMTF8JEUI.PEVMBUJPO
*ODSFNFOUBM&ODPEJOH
0QUP*TPMBUFE.04'&5
'1("
1PXFS4VQQMJFT
t
t
)JHI&GGJDJFODZ1PXFS4VQQMJFT
6OJOUFSSVQUJCMF1PXFS4VQQMJFT
1FSJQIFSBM.PEVMFT
t
t
t
t
t
t
t
t
t
t
t
t
.BTT4UPSBHF
.PUJPO$POUSPM
4ZODISP3FTPMWFS
7JEFP$POUSPM
'JSF8JSF
64#64#
$"/#VT
$"/4QJEFS
(JHBCJU&UIFSOFU
(14
(4.(134&%(&.PEFN
8JSFMFTT5FMFNBUJDT
XXXSUEDPNrTBMFT!SUEDPN
#VT4USVDUVSFT
t
t
t
t
t
1$*F
1$*&YQSFTT
1$*
1$Plus
1$
The products above are just a sampling of RTD’s board-level and
ruggedized modular packaging solutions. From low-power to high
performance, RTD can tailor a system for your mission-critical
application. Visit www.rtd.com to see our complete product list.
AS9100 & ISO9001 Certified
35%&NCFEEFE5FDIOPMPHJFT*OD
Tech In Systems
Embedding Excellence
MSC Embedded offers a large
variety of COM products in COM
ExpressTM, QsevenTM and ETX®.
Our COM products are scalable,
easy to install and powerful to
enable long living industrial applications with an upgrade path
to future needs.
COM Express™
MSC CXB-6S
Intel® CoreTM – 2nd Generation
Multiple i5, i7 and Celeron
processor options available
Single, dual and
quad core solutions
Intel® HD Graphics 2000 / 3000
Intel® QM67 or HM65 Platform
Controller Hub
Up to 16GB DDR3 SDRAM,
dual channel
Four SATA-300 interfaces
One PATA interface
LVDS (24 Bit, dual channel)
and VGA
Resolution up to 2560 x 1600
Five PCI Express™ x1 lanes
Eight USB 2.0 interfaces
V-2_2012-WOEI-5945_RTC
COM Express Type 2
125 mm x 95 mm (4.92 x 3.74”)
36
Untitled-6 1
MSC Embedded Inc.
direct: (650) 616 4068
[email protected]
www.mscembedded.com
3/1/12 9:52:51 AM
additional features that support two different models of programming for parallelism: task parallelism and data parallelism.
Task parallelism is a model with
which embedded engineers are most familiar. Task parallelism is commonly
achieved with a multithreading OS, and
leveraged so that different threads of execution can operate at the same time.
When threads need to access common
resources, mutexes, semaphores, or other
types of locking mechanisms are used.
OpenCL supports this model of programming but it is not its greatest strength.
Data parallelism is used in algorithms
that use the same operation across many
sets of data. In a data-parallel model, one
type of operation, such as a box filter, can
be parallelized such that the same microalgorithm can be run multiple times in
parallel, but each instantiation of this algorithm operates on its own subset of the
data—hence the data is parallelized. This
is the model of programming that OpenCL
is best suited to support. Five compatible
and intersecting models of OpenCL will
help explain the concepts it embodies.
These are framework, platform, execution, memory and programming.
The OpenCL framework consists of a
platform layer, a runtime and a compiler.
The platform allows a host program to
query available devices and to create contexts. The runtime allows a host program
to manipulate contexts. The compiler creates program executables and is based
on a subset of C99 with some additional
language features to support parallel programming. In order for silicon vendors to
provide OpenCL conformance, they need
to provide an OpenCL driver that enables
the framework to operate.
The platform is defined by a host that
is connected to one or more devices, for
example, a GPU. Each device is divided
into one or more compute units, i.e., cores.
Each compute unit is divided into one or
more processing elements.
Execution within an OpenCL program occurs in two places: kernels that
execute on devices—most commonly
GPUs—and a host program that executes
on a host device—most commonly a CPU.
To understand the execution model,
it’s best to focus on how kernels execute.
When a kernel is scheduled for execution
by the host, an index space is defined.
An instance (work item) of the kernel executes for each item in this index space.
In OpenCL, the index space is represented by something called an NDRange.
An NDRange is a 1-, 2- or 3-dimensional index space. A graphical representation of an
NDRange is shown in Figure 1. The host defines a context for the kernels to use. A context includes a list of devices, kernels, source
code and memory objects. The context originates and is maintained by the host. Additionally, the host creates a data structure
using the OpenCL API called a commandqueue. The host, via the command-queue,
schedules kernels to be executed on devices.
Commands that can be placed in the
command-queue include kernel execution
commands, memory management commands and synchronization commands.
The latter are used for constraining the
order of execution of other commands. By
placing commands in OpenCL commandqueues, the runtime then manages scheduling those commands to completion in
parallel on devices within the system.
Work items executing a kernel have
access to the following types of memory:
• Global memory—available to all work
items in all work groups.
• Constant memory—initialized by the
host, this memory remains constant
through the life of the kernel.
• Local memory—memory shared by a
work group.
• Private memory—memory private to a
single work item.
As already mentioned, OpenCL supports two main types of programming
models: data-parallel where each processor
performs the same task on different pieces
of distributed data; and task-parallel where
multiple tasks operate on a common set of
data. In any type of parallel programming,
synchronization between parallel threads
of execution must be considered. OpenCL
offers three main ways to control synchronization between parallel processing
activities. First, there are barriers to constrain certain work items within an index
space to operate in sequence. Second, there
are barriers to constrain the order of commands within the command-queue. And
finally there are events generated by commands within the command-queue. These
tech in systems
events can be responded to in a way that
enforces sequential operation.
Using tools like OpenCL is great for
photo/video editing applications, AI systems, modeling frameworks, game physics, Hollywood rendering and augmented
reality, to name a few. However, there is
also an embedded profile for OpenCL defined in the specification that consists of a
subset of the full OpenCL specification,
targeted at embedded mobile devices.
Here are some highlights of what the
OpenCL Embedded Profile contains:
• 64-bit integers are optional
• Support for 3D images is optional
• Relaxation of rounding rules for floating point calculations
• Precision of conversions on an embedded device is clarified
• Built-in atomic functions are optional
Looking forward, the OpenCL roadmap contains a number of initiatives to
take it to the next level of relevance.
High-Level Model (OpenCL-HLM):
OpenCL is currently exploring ways to
unify device and host execution environments via language constructs so that it
is easier to use OpenCL. The hope is that
by doing this, OpenCL will become even
more widely adopted.
Long-Term Core Roadmap: OpenCL
is continuing to look at ways to enhance the
memory and execution models to take advantage of emerging hardware capabilities.
Also, there are efforts underway to make
the task-parallel programming model more
robust with better synchronization tools
within the OpenCL environment.
WebCL: OpenCL has a vision to
bring parallel computation to the web via
Javascript bindings.
Standard Parallel Intermediate Representation (OpenCL-SPIR): OpenCL
wants to get out of the business of creating compilers and tools and language
bindings. By creating a standardized intermediate representation, bindings to
new languages can be created by people
outside of the OpenCL core team, enabling broader adoption and allowing the
OpenCL intermediate representation to be
a target of any compiler in existence, now
or in the future.
OpenCL has a bright future, but it has
some hurdles to overcome, many of which
are being addressed by current initiatives
within the working group. In the next decade of computing as we continue to see
a proliferation of processing cores, both
homogeneous CPUs and heterogeneous
CPUs/GPUs, we will continue to have increasing needs for sophisticated software
frameworks that help us take advantage of
all of the hardware computing power that
is available on our systems. As this trend
continues, OpenCL is positioned strongly
as an open, free and maturing standard
that has strong industry support and a
bright future.
LogicPD
Eden Prairie, MN.
(952) 941-8071.
[www.logicpd.com].
Khronos Group
[www.khronos.org].
THEMIS TACTICAL SYSTEMS
REDUCING RISK AND INCREASING SURVIVABILITY
UNDER EXTREME ENVIRONMENTAL CONDITIONS
VITA-74 NANOATR
FEATURES
Intel Atom N455 @ 1.66 GHz
1 GB @ 667 MHz DDR3
Electrical per VITA-46 3U VPX
Electrical per VITA-65 OpenVPX
BP Connectors per VITA-57 FMC
4 Slot + Storage
Conduction Cooled with Fins
Dimensions (W x H x D) 4.88” x 4.12” x 4.38”
4.5 lbs (average)
Conduction Cooled
Operating Temberature 40° C to + 71° C
+28 VDC (18 to 36 VDC)
MIL-STD-810G, MIL-STD-461F
TARGET APPLICATIONS
Mission Computing
Payload Control
Real Time Control
Data Recording
Small Storage and Communications Systems
Mobile Robotics
NANOPAK
FEATURES
Intel Atom N455 @ 1.66 GHz
1 GB @ 667 MHz DDR3
VITA-74 Derivative
I/O Through Front Panel Connector
Dimensions (H x W x D) 89 mm X 21 mm X 90 mm
Conduction Cooled
Operating Temperature -40° C to + 71° C
MIL-STD-810G, MIL-STD-461F
TARGET APPLICATIONS
Real Time Control
Data Recorders
Small Storage and Communications Systems
Mobile Robotics
SCAN THIS QR CODE WITH YOUR
MOBILE PHONE FOR MORE INFO
ON THEMIS TACTICAL SYSTEMS
PRODUCTS
BENEFITS
AVAILABLE 3U VPX CARDS
:[H[LVM[OL(Y[;OLYTHSHUK2PUL[PJ+LZPNU
3V^LY7YPJL
:OVY[3LHK;PTLZ
:THSS7YVNYHTZHYL>LSJVTL
;:)*P?<=7?:PUNSL)VHYK*VTW\[LY^P[O0U[LS*VYL;4P*7<
;06*?<=7??4*74**HYYPLY4VK\SL
;:*?<=7?7VY[:(;(:(:9(0+4VK\SL^P[O74*?4*
;:4?<=7?:(;(:(:4HZZ:[VYHNL+YP]L4VK\SL
;.(?<=7?.YHWOPJZ7YVJLZZVY^P[O(4+, .7<
www.themis.com
(510) 252-0870
Untitled-8 1
© 2012 Themis Computer. All rights reserved. Themis Computer, Themis and the Themis logo are
trademarks or registered trademarks of Themis Computer. All other trademarks are the property of their
respective owners.
37
10:20:22 AM
RTC MAGAZINE 3/6/12
JULY 2012
technology in
systems
Parallel Computing with AMD
Fusion-Based Computer-onModules
The integration of powerful graphics processors on the same die with
multicore x86 processors is opening new areas of compute-intensive
embedded applications.
by John Dockstader, congatec
E
mbedded computing tasks are getting
more and more demanding across
all applications. The same applies to
the processors, which must be flexible and
customizable in order to encode or decode
a variety of media signals and data formats
such as JPEG, MP3 and MPEG2.Depending on the specific requirements, a choice
of processor types is available. If the application is highly specific and individual, a
digital signal processor (DSP) is a common
choice. If the application is basic enough
to be handled by an x86 architecture type
processor, the use of a General Purpose
Computing on Graphics Processing Unit
(GPGPU) can enhance performance. AMD
Fusion-based Computer-on-Modules, which
include AMD’s integrated GPGPU, are now
appearing on the market and provide compute capabilities beyond the traditional x86
performance scope (Figure 1).
For a long time CPUs have been required to offer dedicated and often parallel
performance for the processing of complex
algorithms on top of high generic, mostly
serial, processing power. This is necessary,
for instance, when encoding or decoding
high definition video, processing raw data—
such as in industrial image processing—or
performing complex vector calculations in
diagnostic medical imaging procedures. Un-
38
Figure 1
Computer-on-Module Concept
with AMD Fusion.
til now, if processed in an x86 design, these
tasks required high computing performance
with high clock frequencies, resulting in high
energy consumption and heat dissipation.
While multicore technology and continuous
efficiency improvements in processor technology can address these issues to a certain
degree, the fact remains that a speeding up of
the clock rate alone is not enough to meet all
application requirements.
For example, high 3D performance is
required for appealing animation, visualization and smooth playback of HD content. The
graphics core also needs to support the CPU
when decoding HD videos—something that
is of particular importance in medical technology, as in 4D ultrasound or endoscopy, and
also in infotainment applications. The closer
the embedded application gets to the consumer sector, the higher the user expectations.
For this reason, AMD has combined
both technologies in one package with the
release of the embedded G-Series and RSeries platforms. Users can now take advantage of an extremely powerful graphics unit
with highly scalable processor performance.
The so-called Accelerated Processing Unit
(APU) combines the serial processing power
of the processor cores with the parallel processing power of the graphics card. This signals an end to the previous software-based
division of tasks between the processor and
the graphics unit. Simply put, this means the
processor cores can offload parallel tasks
to the graphics unit, thereby increasing the
overall performance of the system far beyond what has previously been possible.
Driven by the consumer market, the performance of graphics cores has steadily increased. In particular, the 3D representation
of virtual worlds has pushed the specialization of graphics cards and created a demand
for high parallel processing capacity. Due to
the variety of graphics data, such as the calculation of texture, volume and 3D modeling
for collision detection and vertex shaders for
Why Should Researching SBCs Be
More Difficult Than Car Shopping?
INTELLIGENTSYSTEMSSOURCE.COM
IS A COMPARISON TOOL FOR DESIGN ENGINEERS
LOOKING FOR CUSTOM AND OFF-THE- SHELF SBCS
AND SYSTEM MODULES.
Today’s systems combine an array of very complex elements from multiple manufactures.
To assist in these complex architectures, ISS has built a simple tool that will source products
from an array of companies for a side by side comparison and provide purchase support.
INTELLIGENTSYSTEMSSOURCE.COM
Tech In Systems
2 DDR3 SO-DIMMS
Memory Controller
SIMD
Engine
Array
VCE
SAMU
Unified
Video
Decoder
x86 Cores
Platform Interfaces
PCI
EXPRESS
x16 PCIe
SATA
LPC
SPI
HDMI
VGA
HD AUDIO
CIR
Unified
Media
Interface
dvi
USB 2
PCIe 4x1
USB 3.0
Optional
Controller Hub
Figure 2
The AMD-R-Series integrates two to four x86 cores along with an SIMD
parallel processing engine originally designed for high-end graphics, but which
can also be used for numerically intensive parallel operations.
geometry calculations, the functions are no
longer firmly cast in hardware, but are freely
programmable. As a consequence, advanced
graphics units provide an enormous and
highly flexible performance potential.
With the help of GPGPUs, this potential can be used not just for the calculation
and representation of graphics, but also
for data processing. Possible uses include
the calculation of 3D ultrasound images in
medical applications, face recognition in the
security sector, industrial image processing
and data encryption or decryption. Certain
types of data—such as from sensors, transducers, transceivers and video cameras—
can be processed faster and more efficiently
with dedicated processing cores than with
40
the generic serial computing power of x86
processors. This is due to the fact that with
a GPGPU it is irrelevant whether the data
generated by the program code is purely virtual or whether it is supplied via an external
source. So it makes good sense to unite the
CPU and GPU in an APU for an even stronger team (Figure 2).
It is not so much the CPU but the APU
performance that is important. This means
OEMs and users need to say goodbye to
the phrase “excellent CPU performance,”
because processing power is no longer defined by the CPU alone. These days the
graphics unit plays a crucial role as well. In
addition to the pure representation of graphics, it is already used in mass applications
such as filtering algorithms of photo editing programs like Photoshop, programs for
encoding and converting video data and
Adobe Flash Player. In the past, developers
struggled with the fact that traditional CPU
architectures and programming tools were
of limited use for vector-oriented data models with parallel multi-threading. With the
introduction of AMD Fusion technology,
that hurdle has been overcome. Easy to use
APIs such as Microsoft DirectCompute or
OpenCL, which are supported by the AMD
Fusion technology, enable application developers to efficiently harness the power of
the graphics core of the APU for a variety of
tasks beyond imaging—provided, of course,
that the graphics core supports it.
The AMD embedded G-Series and RSeries platforms, with integrated graphics, do
exactly this and AMD offers software development kits for it. This makes moving to a new
type of data processing easier than ever before.
In signal processing, a GPU covers
a specific application area. Even though
there are less graphics engines compared
with a DSP processor, the GPU comes up
trumps on programmability. The individual engines can be used flexibly and can be
allocated to different tasks. For example, it
is possible to use 30 engines in parallel for
fast Fourier transform (FFT), 20 engines
for JPG and another 30 for a total of up to
80 possible engines for MPEG2 encoding.
For specific tasks, a GPU is therefore
more efficient than a DSP. In general, applications with less data and simple algorithms are better suited in order to avoid
overloading the system and memory bus.
Good examples from the medical industry are portable ultrasound devices with
low imaging rates or image analysis equipment. Another very exciting application is
the use in multiple security testing processes
to validate the authenticity of banknotes. In
these applications, the developer is not tied
to existing algorithms, but can program his
or her own security mechanisms.
A classic DSP is often used for smaller
applications such as seamless processing of
digital audio or video signals. A distinction
is primarily made between floating and fixed
point DSPs. The DSP is optimized for a single
operation, massively parallelized and achieves
a fast execution speed. Typical applications
include mixing consoles for sound manipulation, hard drives or speaker crossovers.
CPU
System
Memory
Fusion GPU ...
GPU
Memory
Discrete GPU ...
Figure 3
AMD-Fusion-GPU-Architecture. In addition to the integrated GPGPU, an
external graphics processor or DSP can be attached for specialized tasks.
In the future, GPGPUs will be able to
fulfill even more of the classic functions of
DSPs. But it is also clear that a pure DSP application will not be replaced by a GPGPU
(Figure 3). For a GPGPU to perform digital
signal processing effectively, the application
has to support typical computing features.
GPGPUs also work for “simple” embedded computing tasks. AMD Fusion
technology is not exclusively positioned for
specialized applications. On the contrary,
the Computer-on-Module standard COM
Express from congatec with AMD Fusion
can be used across the entire embedded
computing spectrum. Thanks to high scalability—ranging from single core processors to quad core processors based on the
AMD R-Series—the new AMD platform
covers approximately 80% of all application requirements in the embedded market;
from low power right through to high performance applications. Breaking down the
performance spectrum to known standards,
we can also say that the AMD embedded
G-Series platform is scalable for solutions
requiring anything between an Intel Atom
and an Intel Core i5 dual core processor.
It is important to note that this power
calculation does not take into account the
superior graphics performance, which
thanks to the GPGPU can also be used
for other embedded computing tasks. So
depending on the application, the performance potential may even be much higher.
OEMs can therefore implement their entire product range on the basis of a single
processor architecture, regardless of the
specific sector. This not only reduces development time, but also simplifies the supply chain and lifecycle management and
reduces associated costs. For OEMs and
developers who prefer to use core computing components without much design effort and who strive to optimize their supply
chain management by using highly flexible
COTS platforms, Computer-on-Modules
are the appropriate solution.
congatec
San Diego, CA.
(858) 457-2600.
[www.congatec.com].
…www.men.de/safe … www.men.de/safe … www.men.de/safe … www.men.de/safe … www.men.de/safe … www.men.de/safe … www.men.de/safe … www.men.de/safe … www.men.d
tech in systems
Untitled-4 1
Safety-critical
and Redundant
CompactPCI®
and VMEbus
Certiﬁable up to SIL 4
or DAL-A with safe
operating system
N Safe and reliable through
triple redundancy
N Simple software integration
through lockstep architecture
N Voter implemented as IP core
in safe FPGA
N Conductive-cooling options
for harsh environments
N Meets environmental
standards DO-160 and
EN 50155
N Developed according to
DO-254 and EN 50129
N Guaranteed quality based on
ISO 9001, EN 9100 and IRIS
MEN delivers redundant computers
on board and system level
(COTS or customized) for safetycritical control systems for railway,
avionics and industrial.
MEN Micro, Inc.
24 North Main Street
Ambler, PA 19002
Tel: 215.542.9575
www.men.de/safe
41
3/14/11 9:55:04 AM
ECCN.COM
MAKES IT HAPPEN
ECCN.com is the top and most powerful
network for China’s electronic market. It
provides hundreds of thousands of Chinese
engineers with the most up-to-date technology
information from the leading manufacturers and
solution providers from the U.S.A. and Europe.
The all-inclusive platform includes technology
news, new products, a technical library,
webinars, webcasts, training and videos.
Coverage ranges from consumer electronics
to embedded, industrial, automotive and
medical technolgoies. Components supported
range from sensors and analog circuits to
programmable logics and MCU.
Additionally, the portal allows engineers to
purchase products online to help them finish
their prototype and development projects.
ECCN.com has 6 local offices to serve the
engineers and make delivery fast and efficient.
It also has Global Electronics China Magazine
(GEC) to reinforce marketing and promotion.
Buy it online. Buy it in person.
Beijing
Shanghai
Shenzhen
Xian
Wuhan
Hong Kong
ECCN.COM THE CHINA
eCONNECTION
Connecting 2 Continents and over 1,000,000 Chinese Electronic Buyers & Engineers
www.eccn.com
Tel: 010-82888222, 0755-33322333
Fax: 010-82888220, 0755-33322099
technology deployed
Requirements
Engineering Today
Clearly defining and understanding requirements before
embarking on a large and complex software development
project is key to success. The ability to establish and then
fully implement all requirements demands a systematic
method supported by comprehensive tools.
by Marcia Stinson, Visure
R
equirements engineering is the
process of determining user expectations for a new or modified
product. That sounds so simple, doesn’t
it? Of course, those who are in the field
of requirements engineering—also called
requirements analysis—understand the
difficulty and complexity hidden within
that simple statement. Anyone on a development team who has labored to deliver a
project and has discovered that it wasn’t,
after all, what the user really wanted,
understands the difficulty and complexity even more. Requirements engineering
provides a powerful tool for engineering
teams to streamline development, simplify project management, and deliver results on time, on budget, and with the least
possible headaches.
According to a survey by the Standish
Group, about half of all software projects
fail to be delivered on time and within
budget. Even worse, many of those that are
delivered on time are not accepted by the
users and require additional rework. The
most common culprit? Incomplete and
changing requirements. This problem appears to be an even bigger issue than challenges like lack of user involvement and
lack of resources. Especially in complex
projects, there are many layers of requirements that must be understood from the
44
very start of development. Poor requirements management can lead to the development of systems that are expensive, late
to market, and missing key features. Getting the user expectations described correctly is the key to everything else. This is
where requirements engineering can help.
To understand the basic concept of
requirements engineering, consider the
V-model in Figure 1. This model has
been around for years and is still recognized and used today to understand
the concept of a requirements hierarchy
and a related testing hierarchy. Requirements begin at a very high level of understanding and evolve into more and
more technical details. Testing, on the
other hand, begins at the lowest level
and builds up to finally verify that the
system is providing what the users have
asked for and needed.
Formally, requirements engineering
activity is divided into requirements development and requirements management.
Requirements development is composed
of elicitation, analysis, specification, and
validation. Requirements management
is the control of the entire requirements
process, especially handling any change
in requirements. On the other hand, some
practitioners just label the whole activity
as requirements analysis.
Why Is This Important to
Embedded Systems?
In the development of most systems
today, the focus is on functionality. If the
functionality meets the user needs then
there is usually some acceptance on the
part of the users that some of the nonfunctional requirements may not be completely satisfied. In most systems, as long
as the functionality is met, some adjustments to the non-functional requirements
are more readily accepted.
Products are systems that consist of
sub-systems and their interfaces. Each
subsystem is considered an independent
system with interfaces to other subsystems. Each subsystem has its own system
requirements, but because it is associated
with a larger system, it will be subject to
certain restrictions imposed by the system
as a whole.
Embedded systems are subsystems of
the overall system, so their requirements
are derived from system requirements.
The individual component merits its own
requirements analysis, modeling, and further decomposition of requirements, but
these activities cannot be done in isolation—they rely on the requirements they
have been given. The subsystem must
meet these requirements in order to fit effectively into the overall system. . These
Technology deployed
Feasibility Study
/ Concept
Exploration
System Validation Plan
System
Validation
System Verification Plan
(System Acceptance)
and
n
nitio
Defi
Detailed
Design
Unit / Device
Test Plan
pos
ition
om
Subsystem
Verification
Rec
ition
High-Level
Design
Subsystem
Verification Plan
(Subsystem Acceptance)
System
Verification &
Deployment
tion
pos
om
Dec
System
Requirements
Retirement /
Replacement
and
Concept of
Operations
Changes
and
Upgrades
Unit/Device
Testing
gra
Lifecycle Processes
Operations
and
Maintenance
Inte
Regional
Architecture(s)
Software / Hardware
Development
Field Installation
Document/Approval
Implementation
Time Line
Development Processes
Figure 1
V-curve of requirements engineering illustrates the relationship between a requirements hierarchy and related testing
hierarchy.
requirements can be classified as functional and non-functional. Functional requirements specify capabilities, functions
or actions that need to be possible with the
system. That is, what the system shall do.
An example of a functional requirement
would be “The missile shall hit a moving
target,” or “The ground station shall be
able to establish the missile’s target.”
Non-functional requirements specify
the qualities that the product or its functions must have. This means that nonfunctional requirements not only apply
constraints to the system to be built but
also establish its quality and actually play
a key part in the development. An example of quality aspects of a function would
be “The missile shall hit a moving target
in the range of 100 miles,” or, “The missile shall hit a moving target in the range
of 4,000 miles.” Thus, a single non-functional aspect may not only influence the
complete development of a system, but
every constraint also has an associated
cost. Therefore, requirements containing
quality aspects are categorized as nonfunctional requirements, whereas requirements describing capabilities are categorized as functional requirements.
Non-functional requirements that are
allocated to embedded systems normally
impact the entire product, including how
the system is developed and maintained.
Embedded systems must be designed
within the context of the environment in
which they will be running since the environment imposes restrictions on how the
product will behave. In embedded system
development, there must be an increased
focus on handling and recovering from
errors that may occur. Other typical nonfunctional software requirements for embedded systems include synchronous vs.
non-synchronous execution, safety and
reliability, resource constraints and autonomy or the ability to operate without
human intervention.
While the functionalities are normally prioritized by means of costbenefit, constraints are divided into
two types of non-functional requirements: hard non-functional and soft
non-functional. Hard non-functional
requirements are non-negotiable and
must be met by the system or the system is considered to be failed. Soft
non-functional requirements are negotiable with the customer and failure
to fulfill these requirements will not
necessarily cause the system to fail
testing. In embedded systems, most of
the non-functional requirements are
hard requirements. There may be some
room for negotiation but normally that
requires a series of trade-offs with
other components of the system, not
45
technology deployed
Testing Requirements at Various Levels
The missle shall hit
the target.
The missle shall
update its position
every 30 seconds.
The altimeter shall
read terrain data
in 1 second.
verifies
verifies
verifies
Fire a test missle
at a target.
Test in lab to ensure
missile updates
position every 30
seconds.
Test that altimeter
reads terrain data
in 1 second.
Figure 2
To ensure a successful project, requirements must be tested at various levels.
with the end users. Usually the overall
system requirement is a hard requirement that has to be met.
Consider a situation for a control
system that is embedded in a cruise missile. The missile has to read terrain data,
process the data, and update the missile’s
course. The quality aspect of this capability is to do so in 30 seconds. The 30
second limit is a hard non-functional requirement. There is no wiggle room in
meeting it—if the requirement is not met,
the missile will likely miss an update,
become lost and miss the target. On the
other hand, if we establish that the update
frequency shall be every millisecond, the
cost of the system might become unaffordable, or not feasible with the current
technology—and not add any value to the
system compared to the 30 second update.
Knowing this will change the way
developers will design the software.
They may break the 30 seconds down
into even more detailed processes, allocating a portion of the time to each
process. There will be 10 seconds to
read the terrain data, 15 seconds to process the data, and 10 seconds to update
the missile, for example So there may
be some negotiating among the three
processes on how the 30 seconds is allocated. But in the end, the update must be
completed in 30 seconds in order to keep
the missile on track. These requirements
may change the way developers look at
designing and implementing the code re-
46
quired to make this happen. Getting the
real requirements defined right and allocated to the individual components is
a key factor and requires solid requirements engineering practices.
Collecting and Managing
Requirements
To develop a successful project, collecting and managing requirements must
begin on day one. It starts with the request
that initiates the project and continues
throughout deployment. During the initial
phase of a project, requirements gathering begins as requirements analysts communicate with users and spend the time
required to understand the problem they
are trying to solve. Understanding the
problem correctly is essential to building
something that will meet the needs of the
end-user. Misunderstanding here will result in a project that does not function as
desired from the perspective of the users.
User requirements should be documented, along with business requirements
(business rules) and nonfunctional requirements. Key skills to these activities
are good listening skills, communication
skills, and concerted effort to understand
the system from the user’s perspective.
These requirements are often modeled in
a use case structure to help understand the
flow of the requirements. Requirements
must be analyzed for completeness, correctness, and feasibility. As a simple example of this, there may be a requirement
for a missile to read terrain and update its
location within two seconds and the updates will occur every 30 seconds. These
requirements are probably based on analysis that indicates this is the minimum
time required to keep the missile on track.
Obviously this is a requirement that must
be met and this will drive the development
effort to focus on these kinds of performance requirements.
Once the user requirements are defined, the process begins to detail the
steps the system must provide to meet the
user needs. The system requirements are
derived from the user requirements. This
requires a much more technical analysis
and an understanding of the system that
is to be built, as well as other systems that
may interact with the system being built.
The requirements may be developed
in phases. Some may be reused for several
different systems. These kinds of situations lead to a very complex requirements
model that must be thought through carefully. Attributes are usually applied to the
requirements to help understand the variants and releases of the systems in which
they are used. If we refer to the missile
example once again, let’s assume that a
new missile is built that weighs less and
goes even faster. This results in the need
to update the location within one second
and the updates need to occur every 15
seconds. This will be a variant of the missile that will have different requirements
associated with it.
After system requirements are developed, they must be allocated to components of the system. This is typically
where an embedded system picks up requirements. For example, the first missile
will go through three steps to update its
location. Each step will have a time associated with it and the total allocation
of the time will be two seconds or less.
First, it must read the terrain data which
is a requirement for the altimeter. Then
the missile stores the terrain data. The
mission software will then store the data
and compare the data to the onboard map
to determine if an update is required.
The missile must then perform the update. Each step will have a performance
requirement associated with it, the total
of which is no more than the required
three seconds.
Technology deployed
Table 1 shows a sample requirements
traceability matrix (RTM) for a few requirements in a system. Now multiply
that matrix by hundreds and imagine the
complexity of managing that traceability.
Without a requirements management tool
to manage these relationships throughout the system, it is impossible to understand their effect upon one another and
the impact of any changes. Think about
the traceability relationships required to
manage a complex system that has several
layers of requirements (i.e. many subsystems). Traceability ensures that all user
needs are being satisfied by the system
and that no system features are added at
a phase in the project that are not derived
from a user need. Traceability also allows
engineers to see what requirements might
be impacted by a proposed change. If you
consider not only the traceability, but also
attributes (additional information for requirements) that must be maintained, the
task grows even more complex.
Requirements and Other
Development Activities
Requirements form the basis of the
project information. Everything done on
the project is based on the requirements.
Models provide a useful tool for design
teams, enabling them to better understand
the system and develop requirements. A
model is a representation of something
real, produced to represent at least one
aspect of an entity or system to be investigated by reducing complexity. Models
may also be produced for testing ideas. A
model may take a physical form or may be
expressed in the form of a specification.
Select the type of modeling you are
going to do on the project carefully. Keep
in mind that a single model does not represent all aspects of the project, but usually just a single view. Usually multiple
models are used to describe very complex systems and each model has a specific purpose.
Software developers rely on requirements engineers to give them the information they need to design and develop
the system. Collaboration with the developers early in the process will help
ensure technical issues are resolved as
early as possible and then potential solutions are vetted with the users. In most
Requirements Traceability Matrix (RTM)
Business User
Requirement
BR1.
System
Requirement
UR1.
Design
Requirement
Document
SR1.
DD1.
SR2.
DD1.
SR3.
DD2.
UR2.
SR4.
DD3.
UR3.
SR5.
DD3.
SR1.
DD1.
table 1
Requirements traceability matrix shows the complexity and
interconnectedness of the process.
organizations, there is a gap in traceability between system requirements and the
design models. It is important to trace
requirements to design (textual or models) to ensure all requirements are being
considered in the solution. This is also
necessary to do a comprehensive impact
analysis when changes are proposed to
the requirements. Although this takes
time and effort, the benefits of maintaining some kind of traceability to design
artifacts is important.
Requirements also form the basis for
all testing that must take place. Part of the
requirements engineering responsibility is
to ensure that all requirements are tested
and that the system has passed these tests.
In a traceability model, all requirements
should link to some kind of test to ensure
that all requirements are evaluated at the
correct level (Figure 2).
As you can see from this example,
testing at the lowest level can be done
quickly. The altimeter is tested outside
of the missile for performance. The next
level of testing is done in a laboratory environment that simulates the missile’s activity. The final test, actually shooting a
missile is very costly and difficult to do.
The hope is that by testing individual subsystems first the majority of bugs will be
found at this level and not at the end when
the missile is actually launched.
Requirements traceability provides
an essential tool to ensure project success,
not just in terms of performance but also
in meeting time-to-market goals. With requirements traceability, everything on the
project is traced back to a requirement. It
ensures that there aren’t any unnecessary
efforts. If the argument arises on the project about who is doing what, or why, or on
whose direction, the team can always return to the requirements. Without a clear
set of requirements to drive the direction
of a project, we are like a ship floating in
the ocean with no course.
Visure
Montreal, Quebec.
(514) 944-0154.
[www.visuresolutions.com].
47
technology deployed
Out of the Passenger’s
Seat: Requirements
Traceability to
Drive the Software
Development Process
Static requirements traceability has its strong and weak
points. Dynamic traceability can serve as a way to adapt
the requirements traceability matrix as its component
parts change. Not only can the RTM be shown as a way
to connect many parts of a whole, but also to drive the
development process itself.
by Jared Fry, LDRA Technology
R
equirements traceability has become
ubiquitous in the software development process. While useful in many
environments, it is especially so in regard
to safety-critical systems. The requirements
traceability matrix (RTM) is a crucial artifact for the verification process and provides
insight into the many interconnected aspects
of the development ecosystem. An RTM attempts to represent a complex, dynamic environment within a static context. This conversion introduces weakness into the traceability.
Static Requirements Traceability
and its Drawbacks
During the software development process, many artifacts are generated with various links to one another. These artifacts
are wide-ranging in scope. Everything
from high- and low-level requirements, to
models, designs, source code and test cases
may be linked. This connection between
artifacts gives a view into how each object
48
is decomposed into the others. A high-level
requirement may be represented by several
lower-level requirements, each with several models, multiple lines of source code,
and numerous tests associated with them.
These “many-to-many” relationships create a complex web of bonds that can be
difficult to grasp (Figure 1).
The use of an RTM can help to unravel
that web. The matrix provides an organized
structure in which these complex relationships
can be understood. From any point within the
RTM, developers should be able to determine
what it took to get to that point and where they
can go from that point. An example of this is
the verification of safety-critical systems. The
need arises to prove that requirements have
been implemented and behave properly. The
RTM can show where high-level requirements decompose into lower-level requirements, source code and associated tests. Only
when these tests are completed and passed
can the requirement be considered fulfilled.
Despite the many benefits of requirements traceability, it does have some drawbacks, the majority of which stem from a
limited ability to represent the RTM in a sufficiently useful way. Often the matrix is kept
as an addition to the end of each document
or as a spreadsheet. This flat representation
must be constantly maintained to stay up to
date with changing artifacts. Due to the complex nature of what the RTM represents, this
maintenance must be performed diligently.
Unfortunately, this is not always the case.
The RTM can easily be outpaced by the
development of its component parts. This will
usually have the effect of rendering the RTM
incorrect or completely obsolete. Updates to
the traceability are often held until the time
of verification or artifact release. This delay
introduces the potential of traceability errors. Meanwhile, the sheer complexity of the
RTM makes finding and discovering errors
more difficult. The challenge of finding a
single requirement that is not mapped correctly or a test case that is linked to the wrong
source code is compounded when there are
hundreds or thousands of these artifacts.
Developers often rely on the RTM to determine the cost and risk involved in making
changes to specific objects. If a requirement
is being considered for revision, the impact
will not fall to that artifact alone—all upstream and downstream associations will be
affected. Determining the scope of this impact will require the use of the RTM. If it is
out of date, not maintained properly, or difficult to follow, these estimates may not correlate properly to the real world, causing potential catastrophe to the development process.
Traceability Tools to Create a
Dynamic Matrix
Modern requirements traceability
tools can help to alleviate some of these
weaknesses. These tools have been designed to access each artifact of the development process. This includes contact with
other tools such as requirements management tools, modeling tools, and even testing tools. Traceability can then be created
along these objects to form the RTM. Often
this traceability can exist only within the
tool and not require any type of modifications to the objects themselves. This link-
technology deployed
Requirements
Development
Verification
System Requirements
SW Architecture / Design
Test Procedures
Test Cases
High-Level Requirements
Source Code
Test Results
Low-Level Requirements
Executable Object Code
Review and
Analysis Results
Figure 1
The complicated relationships between development artifacts can make a
requirements traceability matrix (RTM) difficult to understand and maintain.
Figure 2
Bidirectional traceability within a requirements traceability matrix is dynamic
and immediately indicates the upstream and downstream impact of changes
within the software development lifecycle. TBmanager is the component in the
LDRA tool suite offering this capability.
ing provides an easily accessible version of
the RTM that replaces the previous means
of storing the RTM as a static object.
Ease of access is not the only advantage,
though. Once the RTM has been developed,
traceability tools provide some powerful
features. A graphical representation of the
traceability can be generated that greatly reduces the complexity inherent in the matrix.
The “many too many” relationships can be
quickly viewed and understood rather than
remaining abstract to the user. Making the
RTM easier to understand and view is a
significant advantage for users attempting
to implement requirements traceability in a
dynamic way. An easier-to-digest RTM im-
50
Untitled-2 1
6/27/12 10:16 AM
proves the chances of finding errors in the
linkage between objects. In fact, many tools
will indicate to the user potential problems
with the traceability, including unlinked objects and incorrect or outdated links.
As the complexity of traceability becomes more transparent, the task of assessing the impact of changes is simplified. The
upstream and downstream impact of a modification can now be quickly evaluated. If a requirement is to be updated, a simple analysis
of the RTM can demonstrate its influence:
• Decomposed requirements or models
may need to be updated.
• Source code that implements the requirement may need to be re-written.
Technology deployed
Software
Requirements
& Defect
Reports
Project
Managers
Manage
requirements;
assign verification &
debug tasks
Map
requirements to
design and
source code
Requirements
Traceability
Matrix
(RTM)
Test Cases
Test
Engineers
Model or
Design
Specification
Software
Engineers
Code Base
Verifies
requirements
against test cases
Implement
requirements &
verify design
Development
& Build
Engineers
Figure 3
User roles enable verification tasks to be distributed appropriately and
managed through the RTM.
• Tests that were developed to verify the
old requirement may no longer be valid.
• Cost and risk can be applied to these
potential changes to create a true estimate that better mirrors the real world.
When linked objects are modified, the
dynamic aspects of modern traceability
tools are brought into focus. The traceability
utility connects to artifacts external to itself.
It can detect when changes have been made
to those artifacts and report to the user. Potential impacts from these changes can immediately be investigated and acted upon appropriately. This can be as simple as re-running test cases to verify that updated code
still meets the linked requirement. It may
also become complex enough to touch each
linked artifact connected to the changed object. Regardless of complexity, a dynamic
RTM can provide the necessary information
in real time as changes are made. This alleviates the problem of the RTM becoming
outdated or obsolete; instead, the most current version of each object will be represented in the traceability. Any impact from
any changes detected will be made visible to
the user, who can take the actions needed to
keep the RTM in balance (Figure 2).
Traceability Matrix as a Driver of
the Development Process
The next step in the evolution of re-
quirements traceability is for it to grow from
a simple artifact of the development process
into the main driver of the process itself. The
foundation for this expansion has already
been laid. The RTM consists of all major
artifacts generated by the software development process and the relationships between
them. Once these links are identified, the associated tasks can be identified as well.
The addition of these tasks to the RTM
is the starting point of the transformation.
Cutting-edge requirements tools are already implementing the necessary features
to allow the RTM to reach its highest potential. For example, each implementation
of a requirement in source code will require analysis and testing. These will generally be expanded on in a software testing
plan that is included as an artifact in the
RTM. Each test outlined in this document
becomes a task that must be completed to
validate the associated requirements. These
tests can then be identified when validation
of the requirement is needed such as when
linked source code has been modified. The
ability of the traceability tools to connect
with the software testing tools provides a
simple and automated way to execute or
update test cases when needed, and to access the results of those tests.
The concept of users and roles is one
of great importance. A user can be defined
and assigned a role within the development
Untitled-9 1
OceanServer Digital
Compass Products:
• Low Cost Family
of Electronic
Compasses
• Solid State
Package
• High Accuracy
Offered in Serial,
USB or TTL
• Under $200.00 in
Low Volume
• Hard & Soft Iron
Calibration
• Fully Engineered
Solution For
Embedded
Applications
ocean-server.com
(508)-678-0550
51
1/12/10 10:03:31 AM
technology deployed
team. Each role is given certain permissions
with respect to how they may interact with
the traceability (Figure 3). For example, project managers may be given full control of
the RTM and its respective objects and also
have the ability to assign tasks, while system
designers may only be given the option of
updating models, other design artifacts and
their associated tasks. And software developers may only have access to the source code.
Once these roles are established, the as-
signment of tasks can begin. The results of
work performed on these tasks are immediately made available to the traceability managers. This will give an up-to-date view of the
software development progress. As tasks are
completed, progress is being made. If changes
are made within the RTM, then new tasks
may be generated or old tasks repeated to
verify that everything still works as expected.
The generation of tasks within an
RTM is not always a single-step process.
A change to one object in the traceability
may cause a cascade of tasks across several users in multiple roles. Take an update
to a requirement as an example. Updating that single requirement is a task in itself that will often fall to the manager or
technical writer role. Any decomposed requirements may also need to be updated.
Design models that are associated with this
requirement may need to be modified by a
system designer. A software engineer may
need to analyze and potentially re-write
any source code that was implemented for
that requirement. Lastly, a design engineer
may need to update test cases or create new
tests that fully verify that the implementation behaves as expected.
Creating and enforcing a workflow is
needed to manage the many paths that are
involved with the software process. Each
step can be defined and assigned to a user/
role. When a task in the workflow is completed, the next step in the workflow is initiated. This includes informing the assigned
users that they have tasks to complete.
The cycle repeats itself until all required tasks have been accomplished. By
using a requirement traceability tool that can
enforce a workflow into the development environment, each user becomes aware of the
tasks that are assigned to them at a given
time. Managers gain visibility into bottlenecks that slow down the development process and can redistribute tasks accordingly
if users are overloaded with responsibilities.
Requirements traceability has long
been a key element in the software development lifecycle. Despite its inherent importance, it has often been lightly
regarded and occasionally mistreated.
By itself, a static RTM can be viewed as
merely another artifact, one that has the
constant potential to be obsolete or contain hard-to-diagnose errors. However,
buried within the lowly RTM are the
makings of a potent tool. By harnessing
the best in traceability tools, the RTM can
become a dynamic powerhouse that not
only links artifacts, but can enforce teamor project-specific workflow to drive the
entire software development process.
LDRA Technology
San Bruno, CA.
(650) 583-8880.
[www.ldra.com].
52
Untitled-19 1
2/3/12 3:55:28 PM
Learn to make your products
Rugged, Low-Power, Fast, Small
and Easy to Use
at the only conference dedicated entirely to flash memory!
Are you struggling with crucial Solid State Drive (SSD) decisions? Can SSDs resolve your application bottlenecks?
How can you maximize SSD performance? Flash Memory Summit will explore new frontiers in enterprise storage
and help you make the right choices. Summit highlights include:
Pre-Conference Seminar: SSDs—The Fundamentals
Forums and Tutorials
11 Keynote Presentations
FMS Theater
August 21-23, 2012
Santa Clara Convention Center
Santa Clara, CA
REGISTER ONLINE
www.FlashMemorySummit.com
COTS Journal and RTC Magazine readers: Enter code SPGP and receive a $100 discount!
technology deployed
Transforming
Code Analysis with
Visualization
As the volume of code becomes ever larger and more
complex, more efficient methods and tools are needed
to analyze it to find and correct defects. A newly
emerging approach of graphical navigation can help
engineers find their way through the thicket of complex
interdependencies.
by Paul Anderson, GrammaTech
C
ode analysis tools for finding programming defects in large code
bases have proven very popular in
recent years because they are effective at
improving software quality. These tools
work by finding paths through the code
that may trigger risky, undefined, or unwanted behavior. For a large code base,
tools may generate many warnings, so it
is important to be able to process these efficiently. Each report must be inspected by
an engineer to determine if it constitutes
a real problem and whether it should be
corrected. If a fix is proposed, then the engineer will want to know what other parts
of the code may be affected by the change.
The process of inspecting a warning to determine if it warrants action
is known as triage. Some studies have
shown that it takes an average of ten minutes to triage a warning report, but there
is a large variance. Many reports can be
dealt with in a few seconds, but the more
complex ones can take significant effort.
They may involve unusual control flow
along paths that go through several procedures located in different compilation
54
units, and can depend subtly on different
variables. The path may be feasible and
risky in some contexts, but infeasible or
harmless in others. Consequently it can be
tricky and time-consuming for engineers
to fully understand reports.
The process of remediation can be
similarly difficult because a proposed fix
can have wide-ranging and unexpected
consequences. A small change to a single
procedure can potentially affect all functionality that depends on calls to that procedure. To be efficient at deploying the fix,
an engineer will want to understand the
other components of the software that are
most strongly dependent on the change,
so that validation activities can be prioritized to focus on those components first.
The process of identifying the affected
components is sometimes referred to as
append_str
ripple-effect analysis or impact analysis.
To deal with complexity, programs
are usually designed so they can be
thought about at different levels of abstraction, and implemented so that those
levels are apparent in the source code.
This is usually helpful but can sometimes
be misleading because the implementation may diverge from the design and the
boundaries of the clean abstractions may
be violated.
The essence of the issue is that programs can be large complicated beasts
with complicated and subtle dependences
between their components. New tools are
emerging that help engineers penetrate
this fog of complexity. Program visualization tools are proving especially useful
at helping engineers gain insight into the
subtleties of their program. When used
appropriately they can amplify the effectiveness of a code analysis tool.
An important property of these tools
that makes them effective is that the visualization is completely and automatically
generated directly from the code itself.
Thus the engineer can see exactly what is
in the code instead of an idealized representation that may hide too many essential
details. The code can be shown at different levels of abstraction from high-level
modules down through compilation units,
then individual procedures and finally as
the text of the code itself.
Until fairly recently, code visualization tools have been limited in the amount
of information they can display. However two trends have converged to make
it possible to have tools that can show
very large quantities of information, yet
still be responsive to user actions. First,
new techniques have emerged for automatically eliding information depending
on the zoom level. Secondly, powerful
video cards with hardware acceleration
features for rendering detailed scenes
return_append_str
Figure 1
A visualization of the call graph showing the immediate neighbors of the
function in which the bug occurs.
strcpy
Technology deployed
cmd_book
BookPGNReadFromFile
cmd_pgnload
PGNReadFromFile
yylex
append_comment
cmd_book
BookPGNReadFromFile
cmd_pgnload
PGNReadFromFile
yylex
append_str
return_append_str
strcpy
Figure 2
A larger fragment of the call graph. Functions in red are involved in a path that is likely to trigger the defect.
Figure 3
Top-down visualization of the source code for a medium-sized program.
have become ubiquitous, and the tools are
now able to take advantage of this. The
combination of these factors means that
powerful new visualization techniques
are feasible. Let’s look at some examples
of how visualization can be used to help
an engineer interpret the results of a code
analysis tool.
Bottom-Up Visualization
Imagine a static analysis tool has reported a potential buffer overrun. The engineer responsible for triaging this warning must ask the following questions:
• Is the warning a real defect? Static
analysis tools make approximations
that can cause false positives, so it is
important to determine this first.
• Is the bug likely to show up in the
field?
• Some buffer overruns are harmless,
but others may cause crashes or may
be critical security vulnerabilities.
What are the consequences of this bug
being triggered?
• The point at which the buffer overrun
occurs is seldom the exact point where
the programmer erred. The error may
be where the buffer was allocated or
where an index into the buffer was
calculated. Where was the error that
gave rise to this bug?
• How should the defect be fixed?
• Finally, are there other defects like
this in other parts of the code?
These questions are all best answered
by starting from the point where the error occurs and working backward and forward through the low-level components
of the code. Take for example a buffer
overrun found in an open-source project.
The offending code appears in a function
named return_append_str as shown
here:
if (!dest) {
newloc = (char *)
malloc(strlen(s))+1;
strcpy(newloc, s);
return newloc;
}
In this case it is easy to confirm that
this is unquestionably a real bug—the
+1 is in the wrong place (it should be
between the parentheses), so the call to
strcpy on the following line will always
overflow the buffer by two bytes. The
next question is to determine if the defect is likely to show up in the field. Note
that the defect is only triggered if the
true branch of the conditional is taken.
Perhaps this code is never deployed in
an environment where that can happen.
To answer this question, it is important
to consider the many ways in which the
function can be called. This is where a
visualization tool can begin to be helpful. Figure 1 shows a visualization of the
subset of the call graph in the vicinity of
the defect. In this figure, functions on the
left call functions on the right. From this
it can be seen that the only call to return_append_str is from the function
named append_str.
The user can then expand the view
by working backward in the call tree to
show more detail. Once enough detail has
been revealed to understand the context,
it becomes evident that there are several
different ways in which the function containing the bug can be called. The next
question is whether some or all of these
are dangerous. Figure 2 shows how this
can be seen in the visualization.
In this case the user has asked the
analysis engine to determine which of the
55
technology deployed
paths leading to return_append_str
are dangerous. The red path indicates that
the defect is likely to be triggered if that
sequence of calls occurs. From here it is
Top-Down Visualization
possible to show a textual representation
of the call path from which it is easy to
find the point of error and begin to plan
a fix.
Not all code analysis tasks are suited
to a bottom-up approach. Sometimes engineers want to take a high-level view
src
swap.c
search.c
quiesce.c
Quiesce
Search
inlines.h
AddXrayPiece
SwapOff
atak.c
eval.c
AttackTo
Bishop Trapped
iterate.c
High modified
cyclomatic
complexity may
indicate this function
is difficult to test
properly.
leadz
GenAtaks
Iterate
KPK
Evaluate
test.c
LoneKing
inlines.h
ScoreKBNK
leadz
ScoreP
TestEval
TestEvalSpeed
DoubleQR7
AttackFrom
nbits
ScoreDev
cmd.c
FindPins
ScoreK
cmd_show
CTL
hung.c
EvalHung
inlines.h
nbits
Figure 4
A visualization of the call graph where the colorization indicates potentially risky values of a metric.
Allison
>'y'ϭϱϬ>ŽǁͲWƌŽĮůĞ&ĂŶůĞƐƐ^ǇƐƚĞŵǁŝƚŚ
/ŶƚĞů®ƚŽŵΡƵĂůŽƌĞWh
^ŽŌǁĂƌĞŶŐŝŶĞĞƌ
Ăƚ>ŽŐŝĐ^ƵƉƉůǇ
Less than 1.5” tall, this system is DIN-rail and surface mountable
ĨŽƌĚŝƐĐƌĞĞƚŝŶƐƚĂůůĂƟŽŶ͘/ͬKŝŶĐůƵĚĞƐ/ŶƚĞů'ď>E͕,D/͕s'͕
ϮǆŚŝŐŚͲĐƵƌƌĞŶƚh^͕ƐŽůŝĚƐƚĂƚĞƐƚŽƌĂŐĞŽƉƟŽŶƐ͕ĂŶĚtŝͲ&ŝ͘
Embedded & Industrial
Technology
Buy Online
www.logicsupply.com/itx
© 2012 Logic Supply, Inc. All products
56
Untitled-2 1
7/10/12 11:05 AM
Technology deployed
of the code. Large programs can contain
hundreds of thousands of procedures, and
there may be millions of calls between
procedures. Clearly it is infeasible to display all of these at once, so visualization
tool designers have developed representations that summarize that information
when the program is viewed from a high
level, yet allow more detail to be revealed
as the user drills down to lower levels.
From a code analysis point of view, a
common use case is for a manager to see
which high-level modules have the highest
density of warnings, and to be able to drill
down through sub-modules to low-level
components and finally to the code itself.
Figure 3 shows a sequence of screenshots from a visualization tool (this is from
CodeSonar) that demonstrates top-down
visualization. Here the module hierarchy
of the code is derived automatically from
the file and directory structure. The leftmost part shows a fully zoomed-out view
of the program. When zoomed out, the
low-level calling relationships between
procedures are projected onto the higher
levels. As the user zooms in, more details
start to emerge—first subdirectories, then
source files, then individual procedures.
The rightmost part shows how the visualization can lead directly to the textual
representation of the code.
Here the layout of nodes is chosen
automatically by a graph-layout algorithm. Tools usually offer users a choice
of different layout strategies. For the toplevel view, a “cluster” layout where link
direction is indicated by tapered lines, as
in Figure 3, is often the most appropriate.
A left-to-right layout is commonly more
useful when showing a small number of
nodes, such as when operating in a bottom-up mode.
Operations on the elements of these
views can be used to help an engineer plan
a fix to a bug. The user can select a function and with a single command can select all other functions that are transitively
reachable from that function. These will
be in components that may be affected
by the proposed fix, so testing activities
should prioritize those parts first.
Additional data can be overlaid on
the visualization to help users understand
what parts of the code warrant attention.
The warning density metric mentioned
above is appropriate, but standard source
code metrics may be useful too. Figure 4
shows a visualization of part of a small
program where components containing functions with increasing cyclomatic
complexity are highlighted in deeper
shades of red. This helps users quickly see
risky parts of the code.
Visual representations of structures
and relationships are well known to be
helpful for users wishing to gain an under-
Untitled-12 1
standing of complex systems. Tools that
generate visualizations of software systems
are particularly useful when used in conjunction with code analysis tools. Together
they can make the process of improving
software quality much more efficient.
GrammaTech
Davis,CA.
(800) 329-4932.
[www.grammatech.com].
57
9:50:36 AM
RTC MAGAZINE 1/11/12
JULY 2012
products &
TECHNOLOGY
Rugged Router Runs Cisco IOS Software
If you try to pronounce SFFR, it will probably come out “safer,”
which is exactly what the so-named SFFR router from Extreme Engineering Solutions purports to offer: safer, secure, encrypted communications, courtesy of the company’s hardware, running Cisco IOS
IP Routing software. At less than 72 cubic inches and 3.5 pounds, the
SFFR router provides mobile ad hoc networking for military, heavy industry and emergency response, extending the Cisco enterprise infrastructure beyond the reach of traditional fixed-network infrastructure.
The X-ES SFFR router incorporates Cisco IOS IP Routing Software with Cisco Mobile Ready Net capabilities to provide highly secure data, voice and video communications to stationary
and mobile network nodes across both wired and
wireless links. Combining the SFFR with
UHF, VHF, Wi-Fi and other radio platforms enables integrators to create mobile,
FEATURED PRODUCT
Family of 12 Volt Integrated Power IC Solutions
Delivers High Power Density
While many small embedded modules may integrate a range of devices with differing power input requirements, it is desirable to have one
power input to the module itself. This, then, entails regulating the power
inputs to serve the different onboard needs—a job that can become complex and space-hungry. A family of fully integrated 12 volt DC/DC converters is implemented in power MOSFET technology with double the
power density over alternative solutions to address this challenge.
GND
S_OUT MASTER
VOUT
VIN M/S
VOUT
VFB
Feedback &
Compensation
VIN
GND
S_IN
SLAVE
VOUT
VIN M/S
REXT
VFB
OPEN
51.1k
Up to 4x 15A EN23F0 devices can be paralleled for high load applications
wireless, ad hoc networks, without requiring a connection to central infrastructure. This rugged router, available in natural convection-cooled,
conduction-cooled, or forced air-cooled enclosures, can be installed
tomorrow in any vehicle or aircraft and/or deployed in the harshest of
environments.
Unique features of SFFR include the rugged router running Cisco
IOS Release 15.2GC, Cisco Unified Communications Manager Express
(CME) support and Cisco Mobile Ready Net, which allows for mobile
ad hoc networking and Radio Aware Routing (RAR) with Dynamic
Link Exchange Protocol (DLEP). In addition there is integrated threat
control using Cisco IOS Firewall, Cisco IOS Zone-based Firewall,
Cisco IOS Intrusion Prevention System (IPS) and Cisco IOS Content
Filtering. Identity management is supported using authentication, authorization and accounting (AAA) and public key infrastructure.
Hardware features include hardware acceleration and hardware
encryption, four integrated 10/100/1000 Ethernet ports available via
commercial RJ-45, industrial IP66/67, or military D38999 connectors. The mini-system footprint is only 4.88 in. (W) x 1.90 in. (H) x
7.70 in. (L) = 71.4 in3. Included is an integrated MIL-STD-704 28V
DC power supply with MIL-STD-461 E/F EMI filtering and the whole
system meets MIL-STD-810F environmental and MIL-STD-461E
EMI specifications.
Extreme Engineering Solutions, Middleton, WY. (608) 833-1155.
[www.xes-inc.com].
58
According to Enpirion, which is emphasizing its focus on miniaturizing DC/DC power systems in applications such as telecommunication, enterprise, industrial, embedded computing and storage
systems, the addition of the EN2300 family seeks to address this
challenge with a line that includes the EN2340QI 4 amp, EN2360QI
6 amp, EN2390QI 9 amp and EN23F0QI 15 amp devices.
These devices capitalize on Enpirion’s PowerSoC technology, which integrates the controller, power MOSFETs, high frequency input capacitors, compensation network and inductor. The
EN2300 family offers small solution sizes with highly efficient
performance, high reliability and a dramatic reduction in time-tomarket. Customers have already validated these benefits with more
than 50 design wins ahead of the official market release.
Enpirion’s proprietary high-speed transistor structure is implemented in 0.18u LDMOS process and excels at the ability to operate at
high frequency while reducing switching losses. This is consistent with
the focus on driving high-speed, low-loss power MOSFET technology
as the key enabler for delivering the highest efficiency solutions with
leading power density. The EN2300 devices offer compact solution
footprints from 4 amp at 190 mm2 to 15 amp at 308 mm2, which represents up to a sixty percent area reduction versus competing alternatives at comparable performance. The devices support an input voltage
range of 4.5 to 14V and an output voltage range of 0.6 to 5V.
Enpirion’s EN2300 devices are available now. The EN2340QI
is priced at $3.56, the EN2360QI at $4.41, the EN2390QI at $6.60,
and the EN23F0QI at $9.00 in volumes of 1k units.
Enpirion, Hampton, NJ. (908) 894-6000. [www.enpirion.com].
PRODUCTS & TECHNOLOGY
Power Efficient Dual Core Processing in a Ruggedized Chassis System
An ultra-compact, fanless system is designed around the tiny VIA EPIA-P900 Pico-ITX board. The VIA AMOS-3002 from Via Technologies
leverages the digital performance of the combined 1.0 GHz VIA Eden X2 dual core processor and the VIA VX900H media system processor (MSP)
on the VIA EPIA-P900 board. The VIA AMOS-3002 offers a powerful, rugged and HD-ready industrial-class PC that combines 64-bit computing
in an ultra-compact system. The highly integrated, all-in-one VIA VX900H boasts hardware acceleration of the
most demanding codecs, including MPEG-2, WMV9 and H.264, in resolutions up to 1080p across the
latest display connectivity standards, including native HDMI support, for next generation multimediaintensive applications.
The system operates completely fanlessly within a robust chassis measuring 19.7 cm x 10.4 cm x
4.9 cm (WxDxH). The VIA AMOS-3002 has a certified operating temperature of -20 to 60 degrees C,
vibration tolerance of up to 5 Grms and a shock tolerance of up to 50G. The VIA AMOS-3002 is also
available with the VIA EPIA-P830 featuring a 1.0GHz Nano E-Series processor, offering an operating temperature
of -20 to 70 degrees C.
Storage is provided through a Cfast slot for a SATA interface Flash drive while an optional storage subsystem expansion chassis offers support
for a standard 2.5” SATA drive. Comprehensive I/O functions on front and rear panels include two COM ports, six USB 2.0 ports, including two of
which are lockable for increased ruggedization, line-in/out, one DIO port, one VGA and one HDMI port for display connectivity and two GLAN
ports for dual Gigabit networking. Optional Wi-Fi and 3G networking are available through a MiniPCIe expansion slot.
Ad Index
VIA Technologies, Fremont, CA. (510) 683 3300. [www.via.com.tw].
Industry-Standard Modular System Pre-Validated
for Intelligent Digital Signage
A new open pluggable specification (OPS)-compliant modular
solution is designed to make digital signage applications more connected, intelligent and secure, which results in devices that are easier
to install, use and maintain. The KOPS800 from Kontron is based on
the OPS specification created by Intel to help standardize the design
and development of digital signage applications that use LCD, touch
screens or projector display technologies. As an industry-standard system solution that can be docked into any OPS-compliant display, the
KOPS800 simplifies development, reduces implementation costs and
speeds time-to-market of a wide variety of enhanced functionality and
graphics-intensive digital signage that
help in delivering a rich user experience
for information and retail systems that
will be part of the retrofit of discrete
legacy systems worldwide.
The Kontron KOPS800 is based
on the Intel Core i7 processor architecture and the Intel 6 Series HM65 /
QM67 chipset. It features a comprehensive range of
externally accessible I/O including Gigabit Ethernet RJ45, two USB 3.0 ports, a HDMI connector and audio jack. The Kontron
KOPS800 also supports OPS JAE interconnect I/O such as HDMI, DisplayPort and USB 2.0 and 3.0. For added security, it supports Intel vPro
with Intel Active Management Technology and features 802.11 a/b/g/n
Wi-Fi for wireless connectivity. It also offers up to 8 Gbyte of dual
channel DDR3-1600 non-ECC system memory and 80 Gbyte mSATA
HD integrated storage.
The system is pre-validated for use with a Microsoft Windows Embedded OS (such as WES7 Pro 64-bit) and Intel Audience Impression
Metric (Intel AIM Suite) technology based on Anonymous Viewer Analytics (AVA) software. Digital signage systems employing the Kontron
KOPS800 with the AIM Suite and running a content management system (CMS) can simultaneously play high-definition video while gathering valuable viewer demographics without invading their privacy to
push custom-tailored messaging to the target audience, which results in
delivering a rich, immersive user experience that can offer significant
infrastructure cost savings.
Kontron, Poway, CA. (888) 294-4558. [www.kontron.com].
Get Connected with technology and
Mini-ITX Module with Third
Generation
Intel now
Core
companies
providing solutions
Get Connected is a new resource for further exploration
Processors
products, technologies and companies. Whether your goal
A new Mini-ITX form into
factor
utilizes the third generation Intel
is to research the latest datasheet from a company, speak directly
Core processors. The
new
WADE-8013
from American Portwell is dewith an Application Engineer, or jump to a company's technical page, the
signed to provide high
performance
and
flexibility
forinfunctional
expangoal of Get Connected is to put you
touch with the
right resource.
sion, and is suitableWhichever
for applications
in gaming,
kiosk,
digitaltype
signage,
level of service
you require
for whatever
of technology,
Connected
will help you
connect with
companies and products
medical/healthcare,Get
defense
and industrial
automation
andthecontrol.
you are searching
for. proThe third generation
Intel Core
cessors are manufactured
on the latest 22nm process technology and are
the most powerful and energy efficient
CPUs from Intel to date. Portwell has
used this technology to create a series
of products that provide smart security,
Getintelligent
Connected with technology and companies prov
cost saving management and
Get
Connected
performance for industrial platforms. is a new resource for further exploration into pro
datasheet from aiscompany, speak directly with an Application Engine
Furthermore, Portwell’s WADE-8013
in touch with the right resource. Whichever level of service you requir
based on the Intel Q77 Express
chipset
Get Connected will help you connect with the companies and produc
and its third generation Core processors
in an LGA1155 socket,
which have integrated the memory and PCI Express controllers supporting two-channel DDR3 long DIMMs and PCI
Express 3.0 to provide great graphics performance.
The WADE-8013 has an extensive feature set, including SATA
storage specification, up to 6 Gbit/s, on four SATA interface (two SATA
III and two SATA II). It also provides connectors for RAID 0/1/5 and
10 modes, and the latest PCIe 3.0 (one PCI Express x16 slot) to support devices for double speed and bandwidth, which enhances system
performance. The primary interfaces of the WADE-8013 include the
latest USB 3.0 high-speed transmission technology, which supports 10
USB ports (four USB 3.0 ports on rear I/O and six USB 2.0 pin headers
on board), two long-DIMM memory slots for DDR3 SDRAM up to 16
Get Connected with companies and
Gbyte, andproducts
three display
types of VGA, HDMI and DVI-D. In addition,
featured in this section.
the WADE-8013
is
equipped
with dual Gigabit Ethernet connectors and
an mSATA via Mini-PCIe, which provides customers with the ability to
choose flexible devices for their storage-demanding applications.
Products
American Portwell, Fremont, CA. (510) 403-3399. [www.portwell.com].
Get Connected with companies and products featured in this section.
59
Ultra Low Power ARM-Based Embedded Computer
Designed for 3.5” to 12” HMI
/LQX[.HUQHO,QVWDOOHG
0%)/$6+0%5$0
0K]$UP&38
'LJLWDO,2/LQHV
%DVH±7(WKHUQHW
86%DQG6HULDO3RUWV
+DUGZDUH&ORFN&DOHQGDU
:DWFKGRJDQG$XGLR,Q2XW
9'&3RZHU
4W\
3UHORDGHGZLWK'26)ODVK)LOH6\VWHP
0+]&RPSDWLEOH3URFHVVRU
.)ODVK.'5$0
3LQ',36RFNHW
'LJLWDO,2/LQHV
6HULDO3RUWV
&RQVROH'HEXJ3RUW
:DWFKGRJELW7LPHUV
9'&RU93RZHU
4W\
Announced by Artila Electronics, the M-606 is an ARM9 WinCE
6.0 single board computer in a standard 3.5” form factor. It is powered
by a 400 MHz Atmel AT91SAM9G45 ARM9 processor
and equipped with 128 Mbyte DDR2 RAM, 128
Mbyte NAND Flash and 2 Mbyte DataFlash.
The M-606 provides one 10/100 Mbit/s
Ethernet, four USB 2.0 hosts, three
RS-232 ports, one RS-422/485 port,
audio, microSD socket and LCD TTL/
LVDS interface. The advanced internal
133 MHz multi-layer bus matrix and 64 Kbyte
SRAM, which can be configured as a tightly coupled memory (TCM),
sustain the bandwidth required by LCD with resolution up to 1280x860.
The resolution of LCD can be configured by using the LCD configuration utility included in the pre-installed WinCE 6.0. The M-606
can drive 5 VDC and 12 VDC backlight of the LCD with up to 3A
current output and PWM brightness control. The M-606 supports .NET
framework 2.0 and user’s application can be developed by VB .Net, C#
and C/C++. In addition, a remote display control utility that provides a
graphic interface for the remote user to control M-606 is also included
in the WinCE 6.0.
Artila Electronics, New Taipei City, Taiwan. +886.2.86.67.23.40.
[www.artila.com].
ZZZMNPLFURFRPVDOHV#MNPLFURFRP
PC/104-Plus Dual Channel Gigabit Ethernet Module
Untitled-3 1
6/25/12 1:14 PM
VERY COOL PRODUCTS!
INTERNATIO NAL
No matter how you shake it, bake it, or configure it, everyone knows
the reputation, value and endurance of Phoenix solid state and rotating
disk VME products. Leading the way in storage technology for decades,
Phoenix keeps you on the leading edge with very cool products!
We Put the State of Art to Workk
www.phenxint.com t 714-283-4800
PHOENIX INTERNATIONAL IS AS 9100/ISO 9001: 2008 CERTIFIED
60
Untitled-6 1
A two-channel, Gigabit Ethernet LAN module is designed to offer flexible, high-performance networking connectivity for industrial
embedded applications. The PPM-GIGE-2 from WinSystems offers
self-stacking I/O expansion on PC/104, EPIC and EBX SBCs based
on the industry-standard PC/104-Plus form factor. This add-in module
uses standard RJ-45 connectors to plug into
10/100/1000 Mbit/s networks using standard
Category 5 (CAT5) unshielded twisted pair
(UTP) copper cables.
Two Realtek RTL8110s are the
Ethernet controllers used by the PPMGIGE-2. They are supported by a wide
range of operating systems including
Windows, Linux and other x86 realtime operating systems. These onboard
Gigabit Ethernet controllers combine a triple-speed, IEEE
802.3-compliant Media Access Controller (MAC) with a triple-speed
Ethernet transceiver, 32-bit PCI bus controller and embedded memory.
With state of-the-art DSP technology and mixed-mode signal technology, it offers high-speed transmission over CAT 5 UTP cable. Functions such as crossover detection and auto-correction, polarity correction, adaptive equalization, cross-talk cancellation, echo cancellation,
timing recovery and error correction are implemented to provide robust
transmission and reception capability at gigabit data speeds.
The PPM-GIGE-2 requires only +5 volts at 500 mA (2.5W). It will
operate from -40°C to +85°C. The module measures 90 mm x 96 mm
(3.6” x 3.8”) and weighs only 88 grams (3 ounces). WinSystems also
offers a single channel version of this board called the PPM-GIGE-1.
Quantity one pricing for the dual channel PPM-GIGE-2 is $199, and
$149 for the single channel PPM-GIGE-1.
WinSystems, Arlington, TX. (817) 274-7553. [www.winsystems.com].
9/9/11 6:36:24 PM
EBX SBC Gains Power Thanks to Third Generation Intel Core Processor
Powered by a third Generation Intel Core processor, a new EBX single board computer boasts a high performance level, combined with a highspeed PCIe expansion site that enables the integration of complex high-bandwidth functions, such as Digital Signal Processing and video processing. These applications have historically been performed with large chassis-based systems and custom hardware.
The Copperhead from VersaLogic offers dual- or quad-core performance that allows high-end compute-bound and
video-bound applications to now be tackled with just a single embedded computer board, not a set of boards in a
rack. This opens new opportunities for automating high-end applications that need to be more portable, rugged, or lower cost than previous CPU architectures allowed.
Based on the industry-standard EBX format of 5.75 x 8 inches, the Copperhead features onboard data acquisition via sixteen analog inputs, eight analog outputs and sixteen digital I/O lines
and up to 16 Gbyte of DDR3 RAM. System I/O includes dual Gigabit Ethernet with network boot capability, two USB 3.0 ports, ten USB 2.0 ports, four serial ports and HD audio. Dual SATA 3 and SATA
6 interfaces support Intel Rapid Storage Manager with RAID 0, 1, 5, and 10 capabilities (SATA 6 ports
only). Flash storage is provided via an mSATA socket, eUSB interface and a Mini PCIe socket. The Mini
PCIe socket also accommodates plug-in Wi-Fi modems, GPS receivers, MIL-STD-1553, Ethernet channels and other plug-in mini cards. The Copperhead supports an optional TPM (Trusted Platform Module)
chip for applications that require enhanced hardware-level security functions.
The Copperhead offers models that support either PCIe/104 Type 1 or SUMIT expansion. The onboard expansion site provides plug-in access
to a wide variety of expansion modules. The PCIe/104 Type 1 interface includes a PCIe x16 lane to support expansion with extremely high bandwidth devices. The SPX expansion interface provides additional plug-in expansion for low-cost analog, digital and CANbus I/O.
Available in both standard (0° to +60°C) and industrial temperature (-40° to +85°C) versions, the rugged Copperhead boards meet MIL-STD202G specifications for mechanical shock and vibration. The high tolerance +12V power input allows the board to operate with power ranging from
9 to 15 volts. This eliminates expensive precision supplies and makes the Copperhead ideal for automotive applications.
Optional high-reliability Ethernet connectors provide additional ruggedization for use in extremely harsh environments. Thermal solutions
include heat sink, heat sink with fan and heat plate. For extremely high-reliability applications, IPC-A-610 Class 3 versions are available. Copperhead is RoHS compliant
VersaLogic, Eugene, OR. (541) 485-8575. [www.versalogic.com].
Get ARM’d with Cogent
Visit our Website @ WWW.COGCOMP.COM
CSB1724-Armada 300
Ɣ 1.6Ghz 88F6282 Sheeva Core
CSB1726-Armada XP
CSB1730B-Armada 500
Ɣ 1GByte 16-Bit Wide DDR3-800
Ɣ 800Mhz 88AP510 Processor
Ɣ Dual PCIe (x1)
Ɣ 1.33Ghz Quad Core (MV78460)
Ɣ Dual SATA Gen 2
Ɣ Dual PCIe (x1)
Ɣ SATA Gen 2
Designed & Manufactured In house !!
17 Industrial Dr. Smithfield, RI 02917
Untitled-1 1
Ɣ PCIe - One x4 and Two x1 ports
Ɣ Dual SATA Gen 2
24 hr email support !!
[email protected]
61
6/6/12 3:32 PM
RTC MAGAZINE JULY
2012
Wireless Sensors Receive Global Frequencies Support
Wireless sensor hardware is now available from Monnit in an added range of frequencies—
in both 868 MHz and 433 MHz ISM radio frequency bands. These radio frequencies are available in addition to the standard 900 MHz wireless sensor hardware released by Monnit in 2010.
The availability of these additional radio frequencies ensures that wireless
sensors can be used for global applications with the 868 MHz frequency
band being primarily used in Europe, Middle East and Africa (EMEA)
and 433 MHz frequency being used in Asia and South America.
“We have made our entire offering of wireless sensors, gateways and accessories available in 900, 868 and 433 MHz operating frequencies to address the immediate demands of our ever
growing customer base. These additional radio frequencies allow
our sensors to be used worldwide, while ensuring reliable low power
and long range operation” said Brad Walters, founder and CEO of Monnit.
Key features of Monnit wireless sensors include support for the 900, 868 and 433 MHz
wireless frequencies. Cellular, Ethernet and USB gateways are also available. Wireless hardware is optimized for reliable, low power and long range operation and in addition, free
online sensor data storage, configuration, monitoring and alerting are available.
Monnit currently provides 28 different types of wireless sensors used to detect and
monitor functions that are critical to business or personal life, including; temperature, humidity, water, light, access, movement and much more. Monnit’s wireless gateways transmit
data between local sensor networks and the iMonnit online sensor monitoring and notification system. All Monnit wireless sensors include free basic iMonnit online sensor monitoring with SMS text and email alerting.
Monnit, Kaysville, UT. (801) 561-5555 [www.monnit.com].
Secure Hypervisor Offers Increased Endpoint and Server Protection
A new version of a secure hypervisor has been designed to offer military-grade protection for the
latest generation of laptops, desktops, servers and embedded systems, helping protect these connected
devices from malicious cyber attacks. With its small footprint, high performance and flexible virtualization support on the latest generation of Intel multicore processors, LynxSecure 5.1 from LynuxWorks
brings the benefits of virtualization to users that have not had the opportunity to use virtualization before
because of size or security issues. LynxSecure offers a truly secure multi-domain platform empowering
users to have separate virtual machines for browsing and corporate functions on endpoints, and also giving the ability to securely host multiple tenants on a single
blade for cloud and server implementations.
The latest LynxSecure version as demonstrated on
one of the latest generation of endpoints, a Dell XPS13
Ultrabook, showcases new key features such as focusbased processor optimization, a new secure console for
user interaction with the secure platform, and new virtual network support for increased inter-virtual machine communication. LynxSecure 5.1 is also
the first version of LynxSecure that supports the 3rd Generation Intel Core processor family.
LynxSecure provides one of the most flexible secure virtualization solutions for use in Intel
architecture-based embedded and IT computer systems. Designed to maintain the highest levels of
military security and built from the ground up to achieve it, LynxSecure 5.1 now offers an industryleading combination of security with functionality, allowing developers and integrators to use the
latest software and hardware technologies to build complex multi-operating systems (OS)-based
systems. LynxSecure 5.1 offers two types of device virtualization, either direct assignment of physical devices to individual guest OS for maximum security, or secure device sharing across selected
guest OS for maximum functionality in resource-constrained endpoints such as laptops. LynxSecure also offers two OS virtualization schemes: para-virtualized guest OS such as Linux offering maximum performance; and fully virtualized guest OS such as Windows, Solaris, Chromium,
LynxOS-178 and LynxOS-SE OS requiring no changes to the guest OS. Another key performance
feature that LynxSecure offers is the ability to run both fully virtualized and para-virtualized guest
OS that have Symmetric Multi-processing (SMP) capabilities across multiple cores.
PARTITION 0
PARTITION 1
PARTITION 2
PARTITION n
LynxOS-SE
RTOS
WINDOWS 7
WINDOWS XP
SECURE DEVICE
SERVER
DIRECT
DEVICE
ASSIGNMENT
ETHERNET
VIRTUALIZED SHARED DEVICES
LynxSECURE
SEPERATION KERNEL AND EMBEDDED HYPERVISOR
GRAPHICS
USB
SATA
PHYSICAL
DEVICE
ASSIGNMENT
ETHERNET 2
PHYSICAL DEVICES
LynuxWorks, San Jose, CA. (408) 979-3900. [www.lynuxworks.com].
62
Rugged COM Express with Third
Generation Intel Core i7 Supports
USB 3.0 and PCI Express Gen 3
A rugged COM Express module is targeted for airborne and vehicle-mounted military computers and human machine interface
(HMI) applications required to function in
harsh environments. The Express-IBR by
Adlink Technology is a COM Express Type 6
module that supports the quad-core and dualcore third generation Intel Core i7 processors
and Mobile Intel QM77 Express chipset. Following Adlink’s Rugged By Design methodology, the Express-IBR is suitable for use in environments prone to severe shock, vibration,
humidity and extended temperature ranges.
The Express-IBR is powered by a quador dual-core
third generation Intel Core
processor and
provides support for USB
SuperSpeed
3.0, PCI Express (PCIe)
Gen 3, and up to three independent displays.
The COM Express module offers up to 16
Gbyte ECC 1333 MHz DDR3 memory in two
SODIMM sockets. Three Digital Display Interfaces can be independently configured for
DisplayPort, HDMI or DVI. PCIe x16 (Gen3)
can serve for external graphics or general purpose PCIe (optionally configured as 2 x8 or
1 x8 + 2 x4); as well as two SATA 6 Gbit/s,
two SATA 3 Gbit/s, Gigabit Ethernet and eight
USB 2.0 interfaces. The Express-IBR with
dual-core processor is validated for reliable
performance in extended temperatures ranging from 40° to +85°C and features a 50%
thicker printed circuit board (PCB) for high
vibration tolerance.
The Express-IBR is a modular, power
efficient solution for applications running in
space-constrained, extreme rugged environments. It is compatible with the COM Express
COM.0 Revision 2.0 Type 6 pinout, which is
based on the popular Type 2 pinout, but with
legacy functions replaced by Digital Display
Interfaces (DDI), additional PCI Express lanes
and reserved pins for future technologies. The
new Type 6 pinout also supports SuperSpeed
USB 3.0 interface, which was unavailable in
COM.0 Rev. 1.0.
ADLINK Technology, San Jose, CA.
(408) 360-0200. [www.adlinktech.com].
REAL-TIME & EMBEDDED
COMPUTING CONFERENCE
WWW.RTECC.COM
W
COME TO
RTECC
REGISTER
IT’S COMPLIMENTARY!
AND MORE AWESOME
THAN WORK!
COMING TO:
IRVINE, CA ON AUG. 21
SAN DIEGO ON AUG. 23
TAKE A DAY
TO LEARN A
BOUT
THE NEWES
T IID
DEAS IN TH
E
EMBEDDED
INDUSTRY.
CHECK OUT
THE LATEST
DEMOS.
LISTEN TO T
ALKS FROM
THE
EXPERTS.
GET OUT OF
YOUR OFFIC
E.
RETURN WIT
H INSIGHT A
BOUT
THE FUTUR
E OF THE IN
DUSTRY.
PICMG 1.3 SHB Features the Latest 22nm Intel Processors
A full-size PICMG 1.3 system host board (SHB) provides high-performance graphics
and flexible PCI Express expansion, and is suitable for a wide range of applications across
several fields including factory automation, image processing, kiosk, medical and military.
The ROBO-8111VG2AR from American Portwell is based on the third generation Intel
Core processors and the latest Intel Xeon processors manufactured on 22nm process technology with energy efficient architecture.
The board features two-channel DDR3 long DIMMs up to 16 Gbyte and ECC to support the Xeon processor E3-1275v2 and the Xeon processor E3-1225v2. The PCI Express
3.0 from the Xeon processors provides three flexible combinations: one PCIE x16, two PCIE
x8 or one PCIE x8 plus two PCIE x4 lanes for versatile applications. The Xeon processors
on LGA 1155 socket are paired with the Intel C216 chipset.
It is also offered with the third generation Core processor family with an integrated, enhanced
graphics engine, which provides significant 3D performance, up to DirectX 11 and OpenGL 3.1
for a broad range of embedded applications. Supporting optimized
Intel Turbo Boost Technology and Intel Hyper-Threading Technology, the third generation Intel Core processor family provides
higher performance and increases processing efficiency. The
Core processors are paired with the Intel Q77 Express chipset.
The Portwell ROBO-8111VG2AR integrates dual Intel
Gigabit Ethernet LAN chips capable of supporting Intel Active Management Technology
8.0 and also features four SATA ports, which support RAID 0, 1, 5 and 10 modes (two ports
at 6 Gbit/s and two ports at 3 Gbit/s). ROBO-8111VG2AR with two serial ports (one RS-232
and one RS-232/422/485 selectable) supports legacy devices while it also provides one parallel port for traditional factory automation applications. The new feature, USB 3.0, provides
upgraded bandwidth from 480 Mbit/s to 5 Gbit/s, greatly reducing the time for data transfer.
Entry-Level Module for COM
Express Type 2 with New Atom
Dual-Core Processors
An entry-level Type 2 Pin-out COM Express module is available with three variants of
the new Intel Atom dual-core processor generation, which are manufactured in 32nm technology. The conga-CCA from congatec is available
with the Atom N2600 processor with only 3.5W
TDP (1M Cache, 1.6 GHz); the Atom N2800
processor (1M Cache, 1.86 GHz) with 6.5W
TDP; or the Atom D2700 processor (1M Cache,
2.13 GHz) with 10W TDP and up to 4 Gbyte
single-channel DDR3 memory (1066 MHz).
The chipset module, which is based on the Intel
NM10, provides improved memory, graphics
American Portwell, Fremont, CA. (510) 403-3399. [www.portwell.com].
Demand-Response Monitoring Unit to Manage Electrical Loads on
Demand
A compact unit for monitoring energy usage at commercial and industrial facilities such
as factories, warehouses, retail stores and office buildings can be connected to utility meters,
plant equipment and facility systems. The OptoEMU Sensor DR from Opto 22 gathers real-time
energy consumption and demand data. It then delivers that data to enterprise business and control systems and web-based applications for monitoring and analysis. In addition, the OptoEMU
Sensor DR helps businesses take advantage of lucrative demand-response (DR) programs from
their local utilities. In response to a request from the utility to reduce power use, the Sensor DR
can signal electrical equipment to shed load. DR programs can provide revenue to businesses in
three ways: first, from discounts for simply agreeing to shed load; second, from actual reductions
in use; and third, from selling electricity back to the utility or energy provider.
The OptoEMU Sensor DR first gathers energy data from up to two utility meters or submeters that emit a standard pulsing signal. Each pulse emitted corresponds to an amount of energy used, and by counting pulses the OptoEMU
Sensor DR can track the total amount of energy used as well as demand. The
OptoEMU Sensor DR can also receive power usage and other data from a variety of devices using the widely adopted Modbus communication protocol. Using Modbus over an Ethernet or serial network, the sensor can communicate
with devices such as temperature sensors and flow meters, Modbus-enabled current transformers (CTs) and power analyzers, as well as larger facility systems
such as plant equipment, building management systems and HVAC systems.
Once gathered by the OptoEMU Sensor DR, real-time energy data is sent to web-based
“software-as-a-service” (SaaS) energy management applications and enterprise business systems, where it can be viewed and analyzed to develop effective energy management strategies
that reduce costs. The OptoEMU DR is available in two models, one for use on both wireless and
wired Ethernet networks, and one for use on wired Ethernet networks only. Pricing is $1,095 for
the OPTOEMU-SNR-DR1, and $895 for the OPTOEMU-SNR-DR2.
Opto 22, Temecula, CA. (951) 695-3000. [www.opto22.com].
64
and display functionalities plus intelligent performance and greater energy efficiency.
The highlight of the COM Express module
is the graphics performance of the integrated Intel GMA 3650 graphics chip. With a clock rate of
640 MHz it is twice as fast as the GPU of the previous Atom generation. In addition to VGA and
LVDS, it has two digital display interfaces that
can be executed for DisplayPort, HDMI or DVI.
Four PCI Express x1 lanes, two SATA
2.0, eight USB 2.0 and a Gigabit Ethernet interface enable fast and flexible system extensions. Fan control, LPC bus for easy integration of legacy I/O interfaces and Intel High
Definition Audio round off the feature set.
The conga-CCA module is equipped with
the new embedded firmware solution UEFI.
The congatec board controller provides an extensive embedded PC feature set. Independence
from the x86 processor means that functions
such as system monitoring or the I2C bus can
be executed faster and more reliably, even if the
system is in standby mode.A matching evaluation carrier board for COM Express Type 2 is
also available. The conga-CCA is priced starting at less than $225 in evaluation quantities.
congatec, San Diego, CA. (858) 457-2600.
[www.congatec.com].
Two Controller Kits Tap into High-Speed PCI Express Bus
Two new controller kits for the ScanWorks platform for embedded instruments from Asset InterTech can accelerate
test throughput by plugging into the high-speed PCI Express bus in the personal computer where ScanWorks is running.
The new controller kits can apply ScanWorks non-intrusive board tests (NBT) to a circuit board. The PCIe-1000 is a
single-TAP controller used for cost-effective JTAG testing. The PCIe-410 controller kit has a four-port interface pod that
can test as many as four circuit boards simultaneously, reducing test times in high-volume manufacturing applications.
Both the PCIe-1000 and the PCIe-410 controllers can also program memory or logic devices that have already been
soldered to a circuit board.
Both of these controllers take advantage of the high speeds of the PCI Express bus in the PC that is hosting ScanWorks.
That means that ScanWorks will execute faster on the PC. Then, when the tests are applied to the unit under test (UUT), ScanWorks can execute
JTAG tests at the speed of the processor, the FPGA or the boundary-scan devices on the board, up to the maximum speed supported by the controller. An additional advantage to the PCIe-410 controller is that it supports parallel test and programming operations via JTAG. Pricing starts at
$4,995.
ASSET InterTech, Richardson, TX. (888) 694-6250. [www.asset-intertech.com].
3U OpenVPX SBCs Bring 10 Gig Ethernet and PCI
Express 3.0
A pair of third generation 3U OpenVPX single board computers (SBC) supports the latest interface technology based on the third
generation Intel Core i7 processors. The two 3U OpenVPX SBCs from
Kontron have native support for 10 Gigabit Ethernet and PCI Express
3.0 to meet the high bandwidth demands of network centric military,
aerospace and transportation applications. The Kontron VX3042 and
VX3044 are specifically designed to provide the appropriate combination of leading-edge performance, power efficiency and bandwidth to
long-lifecycle applications.
The Kontron VX3042 is based on the 2.2 GHz dual-core Intel Core
i7-3517UE processor with configurable TDP between 14W and 25W. It
offers up to 16 Gbyte soldered ECC DDR3 SDRAM and one XMC site
to enable application-specific customization by populating the XMC slot
with additional specialized XMCs including I/O, field bus and storage
modules. Specifically designed for high-performance embedded computing, the leading-edge Kontron VX3044 integrates
the Intel Core i7-3612QE quad-core processor
with 2.1 GHz and up to 16 Mbyte soldered
ECC DDR3 SDRAM. Combined with its
powerful I/O backbone, multiple Kontron
VX3044 enable HPEC systems with an
unprecedented computing density in the compact 3U form factor.
Common to both SBCs are the comprehensive Ethernet connectivity with 10GBASE-KR, 1000BASE-T and 1000BASEBX, eight lane PCI Express gen 3.0 and x1 PCI Express gen 2.0, 1x USB
3.0 and 4x USB 2.0. Storage media can be connected via 2x SATA 3 and
2x SATA 2, both with RAID 0/1/5/10 support. As an option, onboard
soldered SATA-Flash is available to host OS and application code.
Three DisplayPort interfaces provide the increased graphics power of
the integrated Intel HD Graphics 4000 to three independent monitors.
And with the Kontron Smart Technologies with VXFabric’s ability
to use Ethernet TCP/IP over PCI Express, VXControl for monitoring
and control of critical parameters, and the PBIT system test solution,
OEMs have the ability to simplify and accelerate the development of
optimized and highly reliable applications. Their 100 percent backward
and upward compatible pin-out to all Kontron based 3U VPX SBCs,
enables system upgrades without a redesign of the backplane. OEMs
can also profit on the software side, as they just have to write the code
once and can run it on the complete Kontron OpenVPX product range,
enabling real drop in replacements.
Kontron, Poway, CA. (888) 294-4558. [www.kontron.com].
AdDigital
Index
Small Isolated
I/O Signal Conditioning
Boards
Two new isolated digital signal conditioning and termination
Get Connected
with technology
andat
boards for OEM embedded computer
system designers
are aimed
companies
providing
solutions
applications requiring signal isolation
between
a computer
andnow
its moniGet Connected is
a new
resource for further
toring/control points. The ISM-TRM-RELAY
has
16 independent,
Sin-exploration
into products,
technologies
and companies. Whether
gle Pole Double Throw (SPDT)
relays. The
ISM-TRM-COMBO
(pic- your goal
is to research
the optically
latest datasheet
from ainputs,
company,
speak directly
tured) provides a combination
of eight
isolated
eight
with an Application Engineer, or jump to a company's technical page, the
optically isolated outputs, and eight Form C relays on one board. To
goal of Get Connected is to put you in touch with the right resource.
ensure reliable connection
removal
and for
insertion
Whicheverwith
level easy
of service
you require
whateverwith
type offield
technology,
wiring, industry-standard,
3.5
mm
pluggable
connectors
are
used.
Get Connected will help you connect with the companies and products
you are searching
The ISM-TRM-RELAY
has 16for.
independent Form C relays on board.
There are two signal lines,
Normally Open (NO) and
Normally Closed (NC) plus a Common associated for each relay. The contacts can handle 6A
@ 250 VAC / 24 VDC for applications requiring medium current capacity plus isolation.
The ISM-TRM-COMBO offers three
Getcircuits
Connected
with technology and companies prov
different signal conditioning
for
Get
Connected
is
a
new
resource for further exploration into pro
maximum configuration flexibility. There
datasheet
from
a
company,
speak
directly with an Application Engine
are eight isolated inputs that have the catouch with the right resource. Whichever level of service you requir
pability to support eitherinan
active high or active low
Get Connected will help you connect with the companies and produc
signal from 5 to 30 volts. Every input line is optically isolated from the
others and from the computer interface circuits with an isolation voltage
rating that exceeds 2500V. Then each input line is wired to a contact
bounce eliminator to remove extraneous level changes that result when
interfacing with mechanical contacts such as switches or relays.
There are eight isolated outputs for applications requiring medium
current capacity. Each output has a NPN Darlington transistor pair with
an integral clamp diode for switching inductive loads and transient suppression. The collector-emitter voltage can withstand up to 30 volts
and each output is capable of sinking 500 mA of current. The isolation
voltage rating exceeds 2500V. The ISM-TRM-COMBO also has eight
independent SPDT Form C relays that can handle 6A @ 250 VAC / 24
VDC for applications requiring medium current capacity plus isolation.
Both modules
are RoHS-compliant and can operate over an industrial
products featured in this section.
temperaturewww.rtcmagazine.com/getconnected
range of -40° to +85°C.Special configurations can be populated for OEM applications. Quantity one pricing for the ISM-TRMRELAY is $299 and for the ISM-TRM-COMBO $255.
Products
WinSystems, Arlington, TX. (817) 274-7553. [www.winsystems.com].
65
with an Application Engineer, or jump to a company's technical page, the
goal of Get Connected is to put you in touch with the right resource.
Whichever level of service you require for whatever type of technology,
Get Connected will help you connect with the companies and products
you are searching for.
Advertiser Index
Get Connected with technology and companies providing solutions now
Get Connected is a new resource for further exploration into products, technologies and companies. Whether your goal is to research the latest
datasheet from a company, speak directly with an Application Engineer, or jump to a company's technical page, the goal of Get Connected is to put you
in touch with the right resource. Whichever level of service you require for whatever type of technology,
Get Connected will help you connect with the companies and products you are searching for.
CompanyPage Website
CompanyPage Website
ACCES I/O Products, Inc............................. 7.......................................www.accesio.com
Phoenix International.................................. 60.................................... www.phenxint.com
Advanced Micro Devices, Inc...................... 68.........................www.amd.com/embedded
Phoenix Technologies, Ltd.......................... 20..................................... www.phoenix.com
American Portwell...................................... 21.....................................www.portwell.com End
Real-Time & Embedded
of Article
Products
Calculex..................................................... 67.....................................www.calculex.com
Computing Conference............................... 63......................................... www.rtecc.com
Cogent Computer Systems, Inc................... 61................................... www.cogcomp.com
products featured in this section.
Commell....................................................
50................................www.commell.com.tw
Datakey Electronics.................................... 24................................. www.ruggedrive.com
RTD Embedded Technologies, Inc............ 34, 35..........................................www.rtd.com
Schroff.......................................................
........................................ www.schroff.us
with companies mentioned in this 30.
article.
Super Micro Computer, Inc.......................... 5................................. www.supermicro.com
Dolphin Interconnect Solutions..................... 4...................................www.dolphinics.com
Themis Computer....................................... 37.......................................www.themis.com
ECCN.com.............................................. 42, 43.......................................www.eccn.com
Get
Connected
with companies
mentioned in this article.
USB Modules
& Data
Acquisition Showcase.
... 15.................................................................
Xembedded................................................ 31............................... www.xembedded.com
Elma Electronic, Inc..................................... 2...........................................www.elma.com
Get Connected
Extreme Engineering Solutions, Inc............. 11...................................... www.xes-inc.com
Flash Memory Summit................................ 53.................. www.flashmemorysummit.com
Inforce Computing, Inc............................... 17.......................www.inforcecomputing.com
Innovative Integration.................................. 52........................... www.innovative-dsp.com
Intel Corporation...................................... 18, 19........................................www.intel.com
Intelligent Systems Source.......................... 39............ www.intelligentsystemssource.com
JK Microsystems, Inc.................................. 60...................................... www.jkmicro.com
Logic Supply, Inc........................................ 56................................ www.logicsupply.com
Measurement Computing Corporation......... 14.....................................www.mccdaq.com
MEDS........................................................ 33...............................www.medsevents.com
ARE YOU
Men Micro, Inc........................................... 41.................................. www.menmicro.com
A seasoned embedded technology professional?
Microsemi Corporation............................... 25................................. www.microsemi.com
Experienced in the industrial and
military procurement process?
MSC Embedded, Inc................................... 36...........................www.mscembedded.com
Nallatech.................................................... 57................................... www.nallatech.com
Ocean Server Technology, Inc..................... 51............................. www.ocean-server.com
One Stop Systems, Inc............................... 49.........................www.onestopsystems.com
Interested in a career in writing?
CONTACT SANDRA SILLION AT THE RTC GROUP
TO EXPLORE AN OPPORTUNITY
[email protected]
RTC (Issn#1092-1524) magazine is published monthly at 905 Calle Amanecer, Ste. 250, San Clemente, CA 92673. Periodical postage paid at San Clemente and at additional mailing offices.
POSTMASTER: Send address changes to RTC, 905 Calle Amanecer, Ste. 250, San Clemente, CA 92673.
66
THE OTHER GUY’S
HO-HUM
SOLID STATE RECORDER
T
BT
CALCULEX ‘s
MONSSTR®SOLID STATE RECORDER
ON STEROIDS!!!
TM
Recorder Integrated Processor and Router
For details, contact:
CALCULEX
®
132 W. Las Cruces Avenue, Las Cruces, NM 88001
575-525-0131
www.calculex.com