Saturday, June 4, 2011

Combining FPGAs & Atom x86 CPUs for SBC design flexibility

Field Programmable Gate Array (FPGA) technology has been a useful design resource for quite some time and continues to be a mainstay because it delivers many of the same benefits of x86 processor architectures.

The various common advantages include multifunctionality, a healthy and broad-based ecosystem and a proven installed base of supported applications. Combining x86 processor boards with FPGA-controlled I/Os expands these benefits even further, allowing dedicated I/Os to support a wider range of application requirements.

Employing next-generation x86 processors with FPGAs onto a single hardware platform provides the ability to eliminate chipsets so that different areas of applications can be built on the same platform requiring only the exchange the IP cores.

New x86-based embedded computing platforms combined with FPGAs enable a new realm of applications – providing highly adaptable feature options for designs that have previously been restricted due to lack of interface or I/O support.

By understanding the collective advantages of this approach, designers can reduce Bill of Material (BOM) costs and maintain long-term availability with legacy interfaces and dedicated hardware-based I/O. Further, legacy systems now have a bridge to tap into the latest processor enhancements such as graphics media acceleration, hyperthreading and virtualization for greater success in matching exacting requirements.

This is a significant advancement in bridging newer technologies with older systems implemented in military, industrial automation and manufacturing environments.

Blending x86 and FPGAs for adaptable design options


Most x86 architecture designs are paired with a chipset, usually a two-piece component with a specific set of integrated features. Ethernet, memory control, audio, graphics, and a defined set of input/output interfaces such as PCI, PCI Express, SATA , and LPC are routinely integrated options.

However many of these chipsets are moving away from the legacy interconnects (e.g., IDE and PCI) commonly found in deeply established environments such as military, industrial automation or manufacturing systems for safety and operations.

As a result, these industries have not been able to take advantage of processor advancements and subsequent improvements in power and performance.

The availability of new x86 processors in combination with an FPGA presents an entirely new design approach that virtually take away embedded limitations from a predetermined feature set. These capabilities distinguish the Intel Atom E6x5C processor series as a milestone in x86 technologies, and a departure from using a chipset with a fixed definition.

Instead the Intel Atom E6x5 processor is combined with a flexible and programmable Altera FPGA on a single compact multi-chip module. By incorporating PCI Express rather than the Intel-specific Front Side Bus, the FPGA is connected directly to the processor rather than to a dedicated chipset, resulting in maximum flexibility and long-term design security.

Designers further have ready access to increased performance integrated into smaller form factors that offer very low thermal design power (TDP).

Because the FPGA can be programmed to support modern as well as legacy interfaces, OEMs now have a workable migration path from non-x86 to x86 architectures – enabling slower moving technology-based market applications to progress to next-generation processing technologies.

Cementing this approach as an appealing long-term design solution, Loring Wirbel of FPGA Gurus estimates that the CAGR for FPGAs will continue at a strong 8.6 percent which will put the FPGA market at US$7.5 billion worldwide by 2015.

Information is shared by www.irvs.info

No comments:

Post a Comment