Skip navigation
2008

Recent buzz about the 3G iPhone got me thinking about graphics

and video. The iPhone supports OpenGL ES, which provides a 3D rendering system

that can be hardware accelerated, as well as “amazingly zippy” downloads per Steve

Jobs. Network connectivity with ease of downloading content is going to spur demand

for excellent graphics/video in the consumer iPhone and in embedded systems. As you

would expect the embedded market is going to want to take iPhone graphics capabilities

several steps further. For smaller portable embedded products Kontron’s new KTUS15/mITX,

with Intel® Atom™ and Intel® US15 embedded chipset on a miniITX motherboard, or

nanoETXexpress-SP (55 mm x 84 mm) and microETXexpress-SP

(95x95mm) COM Express™ compatible modules offer Intel® GMA 500 for advanced

integrated 3D graphics and HD video decode. We find that Intel’s

multi-threading solution provides our customers with increased flexibility of a

fully programmable pipeline for easier feature upgrades along with simultaneous

processing of graphics and video threads needed for emerging applications. This

solution ensures excellent graphics/video capability for the smallest of

embedded systems.

 

 

 

 

But is that enough? When I think about graphics and video

for embedded systems I like to ponder gaps that may cause usage model

headaches. For the Intel® System Controller Hub US15W it looks like Intel has

done a good job anticipating the embedded community’s needs. With the fully

optimized hardware acceleration of H.264, MPEG2, VC1 and WMV9 and 4 HD audio

streams, developers are better able to maintain thread synchronization while

designing parallel code to take full advantage of the multi-core processing

power for graphics/video rich applications.

 

 

 

 

Of course with innovation there are always new applications

that may require some additional SDVO assistance from Kontron to push the

graphic/video capabilities in custom solutions. Now that I just paid $4.75/gallon for

premium gas for my SUV, I am anticipating that the auto industry may want to offer

more impressive gas security than alocking cap. How about adding another

Intel Atom camera system with video encodingcapabilities to watch that investment?

 

 

 

 

Kontron – Nancy Pantone

 

 

Message Edited by serenajoy on 03-11-2009 08:43 PM

Multi-core processor performance depends on a software developer's ability to make efficient use of the additional cores. Multi-core processors offer developers more avenues to increase system computing performance. Most high-performance CPUs implement cache memory to improve performance. To lower data latency, the execution unit is surrounded by small pieces of high-speed SRAM. The execution unit can access cache memory about 80 times faster than it can system memory. A packet-processing application that's performing TCP reassembly on a large number of TCP flows is likely to access a large amount of data over many memory locations. This results in a lower locality of reference than an application that's performing TCP reassembly on a smaller number of TCP flows. The transition to multi-core processors promises more than an increase in the number of execution cores and computational capability. It offers additional flexibility for development and optimization of higher performance applications. In particular, multi-core architectures can significantly improve program flow so that cache memory associated with each individual execution core is used more effectively. With multiple caches available to software developers, it's possible to optimize data locality, driving higher cache hit rates and improving overall application performance.

 

Many service providers are pursuing IP-based technology to help them deploy new video and rich media services and generate new revenue streams. This means delivering voice, video, broadband Internet and mobile services, and enriching the experience of customers. Keeping up with changing market demands, equipment makers are providing platforms that seamlessly connect multiple access networks, like PSTN, xDSL, Wi-Fi and corporate LAN, using packet-based technologies that can enable a broad mix of services. Similarly, packet processing is integral to a range of applications such as intrusion detection, VPN, firewall, gateways, routers and storage. Equipment manufacturers are under pressure to deliver systems supporting converged IP-based traffic. When developing any application, engineers need to make choices as to how much overhead their system can handle while producing acceptable system performance. As with any multi-processor system, the predominant challenge is to ensure that all the processors are kept busy doing useful work rather than wasting CPU cycles waiting for another core to release a shared resource

 

 

A new class of multi-core processor has begun to appear in a variety of storage, security, wireless base stations, and networking applications. This new class of multi-core processor is made up of eight, sixteen, or even sixty-four individual processor cores with integrated memory controllers, various I/O interfaces, and separate acceleration engines. The new class of processor has made great strides in overcoming the limitations of earlier generation processors. Some companies that develop these processors add threading capability to overcome memory latency, and also include a native 10 Gb/s interface, while others include security engines and even regular expression engines that support very special applications.

It is easy to see that small form factor solutions are the market trend of the future; they allow embedded solution providers to make their devices smaller and more portable. The new solutions successfully extend the embedded application from the office to any place that is accessible by people. In the past, small form factors were not easy to achieve because higher performance normally comes from bigger chips and higher power consumption. This really inhibited the implementation of small form factor designs.

 

The situation is changing now. The Intel® AtomTM platform has removed the limitations previously described and moved the demand for small form factor systems ahead. Based on Intel's 45 nm processor technology, the size of the processor is only 13 x 14 mm, just about 14\% of the size of other processors. With the Intel® System Controller Hub US15W, all functions are integrated into a single chip, including LVDS, USB, HD Audio, a PCIe interface, etc. This allows engineers to reduce the board size without sacrificing features. Apart from the improvements of chip size and functions, the Intel® AtomTM platform brings another innovation for small form-factor applications: extreme low power consumption, less than 5 watts. The result is not only a smaller board, but also a longer battery life and lower thermal output.

 

 

Advantech is providing customers with an embedded SOM solution based on the Intel® AtomTM platform; a new 95 x 95 mm form factor which saves 24\% on board size compared to original SOM solutions, with nothing sacrificed. It supports 24-bit LVDS, USB 2.0, and a PCIe interface and processor speed of over 1 GHz. It is definitely the best solution following new market trends in small form-factor applications.

Recently, I was asked to comment on the impact of the sub-five watt, 45nm architecture. -- It's incredible! The Intel Atom processors deliver when it comes to performance, low power consumption and especially size. Because we've been witness to shrinking processors and their lowering TDPs, it's foreseeable that this trend soon will be applied to multi-core processors. It's music to the ears of mobile computer designers who've already been quick to adopt dual and quad-core processors for tablet-size solutions.

 

 

 

 

 

The migration path to multi-core, single-hand-held solutions has been blazed. The nanoETXexpress specification offers the wide-range power supply support needed for battery-powered designs and it's compatible with the COM Express Type I pin-out defined by PICMG. Best of all, it's the exact footprint requested time and again by designers to fit in the palm of their hand. Now, as the first fist-held solutions are developing around Intel Atom processor based "Nano" form factor COM Express compatible computer-on-modules (example: nanoETXexpress-SP), future upgrades to multi-core processors will be just that - upgrades, not whole new developments. Imagine the possibilities because this is just the beginning.

 

 

 

 

 

 


Christine Van De Graaf

 

 

Product Marketing Manager, Embedded Modules Division

 

 

Message Edited by serenajoy on 03-11-2009 08:44 PM

At the end of May I had the opportunity to present at the MicroTCA Summit which took place in Chantilly Virginia. This is the premier opportunity for vendors, customers and other interested parties to get together for a few days to review product and adoption progress for the increasingly popular MicroTCA standard. For those who are not familiar with MicroTCA, it is an open standard using mainly postcard sized modules in a rack system. What makes MicroTCA different is that the underlying management and connectivity are derived from the highly robust AdvancedTCA standard and so enjoys an IP centric control architecture with the flexibility to add additional fabric interfaces to suit applications.

 

MicroTCA is a relatively new standard (it was ratified mid 2006) and there are still many opportunities for education. Like all new standards, one key question is whether the necessary building blocks are available and will work easily together – after all one of the key tenets of an open standard is that it allows customers to pick and mix components to get a best in class solution.

 

The highlight for me, and also many visitors judging by the comments I heard was the multi-vendor interop workshop. Consisting of 4 different chassis and a multitude of AMC modules, this was one of the first public demonstrations of the growing confidence and maturity in this standard. All the chassis and modules were managed by the Spiderware M3 product from Emerson Network Power to graphically show the element status and other detailed information whilst the AMC modules were running video streaming applications.

 

As a vendor, we always like to demonstrate our own latest and greatest products and I’m sure we’ll continue to do that. However, we shouldn’t lose sight of the real value to our customers of choosing an open standard architecture like MicroTCA that interoperates correctly and so I expect that this will be a model for future coopetition.

With the completion of the AdvancedTCA standards near the end of 2005, a whole new series of industry standard specifications for developing high speed interconnect technologies, next generations silicones, and improved reliability, manageability and serviceability became available for the next-generation carrier-grade communication equipment.  The physical design of AdvancedTCA was aimed at the telecom carrier end office or switching environment.  It can be space, energy, and capital consuming.  However, thanks to these new standard specifications, a variant of AdvancedTCA form factor called MicroTCA, which may be considered as a small form factor version of the new carrier grade technology, was created. One may say that MicroTCA is a lower-cost and smaller version of AdvancedTCA.  But, it still provides support for requirements such as NEBS, ETSI, and ITU specifications.  Due to this capability and its modularity, scalability, and reliability, MicroTCA systems have not only found their applications such as in WiMax base stations, remote terminals, subscriber loop carrier, digital subscriber loop access multiplexers and IP video, VOIP, and for CATV multiple system operators deploying CATV modems, but they are also starting to find their ways outside of the carrier plant environments such as transportation, industrial, government, and aerospace.  One important element besides the power module and the backplane that make up the MicroTCA carrier is called the MicroTCA Carrier Hub (MCH). It provides the management, clock, and fabric hub signals to each AMC position of the backplane.  In MicroTCA, AdvancedTCA AMC modules can connect directly to the chassis backplane without modifications. They are the primary components of MicroTCA.

 

 

 

Being created to address cost-sensitive and physically smaller applications, MicroTCA may not offer as high capacity and high performance as AdvancedTCA. However, its design intent for mission critical systems and high-availability applications is gaining more and more product inquiries from various markets. And, how small and agile can a MicroTCA system be? Well, considering the 45nm low-voltage processor or the integrated SOC, Tolapai, from Intel, and the small physical size of the single-width AMC PCB, anything reasonable seems accomplishable.

Last week as a Premier Sponsor of the MicroTCA Summit in Chantilly, Virginia, Kontron observed the rising interest levels in MicroTCA among customers across all market spaces. With a varied audience and plenty of industry support, MicroTCA is set to make big waves across all markets, namely Defense, Aerospace and Communications. Kontron is excited to be part of this revolution and is constantly adding new MicroTCA integrated platforms and AdvancedMCs to our portfolio.

 

Kontron recently published another article on MicroTCA which is running in the May edition of RTC magazine. “MicroTCA on the Road to Stardom” is co-authored by David Pursley and Sven Fruedenfeld, both of whom spoke at the Summit last week.

 

What are your thoughts on MicroTCA across the major market spaces?

 

 

 

 

 


Sara Fest

Associate Product Marketing Manager

Message Edited by serenajoy on 03-11-2009 08:22 PM

Filter Blog

By date: By tag: