comment on this article

Taking the platform

Taking the platform

FPGA design has become complex enough to demand a platform based approach.

There was a time when programmable logic was a relatively humble technology. The approach was generally used as a way of fixing board layout errors and became widely regarded as 'glue logic'.
But times change. As process technology has careered towards ever smaller feature sizes, programmable logic developers have produced increasingly capable devices. Where once the asic was king of the electronics system jungle, the programmable logic device – known more broadly as an fpga – is ruling the roost.
Brent Przybus, director of product marketing for Xilinx, noted the change. "FPGAs have moved from being a peripheral component to being the centre of a design."
It's a trend that has been in process for a decade. Since the late 1990s, companies such as Altera and Xilinx have recognised that asics were their target. The reasoning included the fact that asics were becoming increasingly more complex and expensive. Implemented on today's leading edge processes, it might cost some $100million to take an asic to market – not something for the faint hearted.
What works against the asic is the high cost of making a mistake. Respinning an asic design not only costs a lot, it also means the product which it enables gets to market later – and that could have serious financial implications.
While the asic has climbed higher up the complexity curve so too has the fpga. In many respects, today's fpgas could be likened to the asics of a decade ago. But there's one crucial difference – programmability.
In fact, Xilinx believes the industry has reached what it calls a programmable imperative. In other words, Xilinx believes the fpga has become the 'defining integrated circuit' for electronics design.
Responding to this increase in complexity, Xilinx has launched the latest development of its Virtex family. According to the company, Virtex-6 has been developed to meet what it sees as 'insatiable demand' for higher bandwidth and for lower power systems.
At the same time, it has added another 'buzz phrase' to the electronics industry's argot – the targeted design platform. Przybus explained the concept. "We have realised that, while we're having success providing fpga solutions, a recurring theme is that we can't provide everything in silicon. We can't harden many features – because that would make our products assps – so that means we need to provide IP cores."

Author
Graham Pitcher

Related Downloads
17638\P24-25.pdf

Comment on this article


This material is protected by Findlay Media copyright See Terms and Conditions. One-off usage is permitted but bulk copying is not. For multiple copies contact the sales team.

Enjoy this story? People who read this article also read...

What you think about this article:


Add your comments

Name
 
Email
 
Comments
 

Your comments/feedback may be edited prior to publishing. Not all entries will be published.
Please view our Terms and Conditions before leaving a comment.

Related Articles

A companion for ASICs

Lattice Semiconductor has unveiled the ECP5 family of FPGAs, designed for use ...

NI Trend Watch 2014

This report from National Instruments summarises the latest trends in the ...

Capactive sensing

This whitepaper looks at a number of capacitive sensing applications to ...

BEEAs 2013

The sixth British Engineering Excellence Awards (BEEAs) will be held on 9th ...

Better batteries

For much of the last Century, battery technology didn't really need to ...

The tale of the tape

Data storage, at least at the consumer end of the scale, is dominated by flash ...