Skip navigation
2010

I'll admit it: I'm envious of TI's Gene Frantz. Not because of his technical contributions to DSP--well, maybe a little--but mostly because his status as Principal Fellow at TI, an elite group to which few are invited, has made him somewhat bulletproof at his company, a position we would all love to achieve.

 

Here's what I mean. As a preamble to his talk, "What to do with a billion transistors" at the Multicore Expo segment of ESC, Frantz was asked by seior management whether he was going to express his views or those of TI. "Mine," Frantz said, to which management replied "OK but can we at least look at your presentation to see if it is close to our view." Frants replied "Sure" but told the ESC audience that anyone who assumes that his talk represents the postion of TI "is wrong, It might be someday, but isn't now."

 

You've got to love this kind of independence.

 

According to Frantz success in multicore will hinge on a number of factors, including:

 

Moore's Law to determine the number of transistors on a processor and Amdahl's Law to determine the proper number of processor cores

 

Three vectors: greater performance, lower cost and reduced power consumption

 

Overcoming the bottleneck of processor to processor communications

 

Choosing a system design that solves a particular customer problem rather than trying to make the problem fit into a given solution

 

Frantz also reminded the group that multicore is not new. In 1990 TI developed the TMS7000 with 1,000 DSPs. Although they were all 1-bit DSPs the chip worked well in the early days of digital TV, processing each pixel and each line in a progressive scan or picture-in-picture application. Before too long, Frantz reminded everyone, compressed signal technology took hold and the processing tasks could be handled by a single DSP.

If pugilists enter a prizefight with too much respect for one another the spectators are likely to watch an event that resembles fisticuffs less than it does "Dancing with the Stars". That, unfortunately is what happened yesterday during the ESC's "No holds barred panel debate entitled "Standalone vs. Embedded Processors" with Clive Maxfield championing FPGAs and Jim Turley taking up the case of the standalone processor.

 

Amid the good natured jabs at one another and some witty repartee the usual pedestran arguments were made by both sites. To wit:

 

The flexibilty of FPGAs when last minute design changes occur

 

The lower cost of hard processors when a project involves large volume production

 

The greater availability of development tools for processors

 

The faster time to market for FPGAs

 

You get the idea.

 

Unfortunately, only passing reference was made by Mr. Maxfield to a number of recent programmable logiic developments that, while not changing the equation, at least would make the discussion more interesting. Some for instances:

 

Altera is positioning its new Stratix V for applicatioins that now use a DSP chip, such as video processing and wireless base stations. The Stratix V FPGAs have a variable precision DSP block that can be set for 9 to 54 bit precision. There are plenty of DSP blocks--up to 3,680--resulting in a great deal of performance up to a total of 1,840 GMACs.

 

At ESC Actel displayed its SmartFusion mixed signal FPGAs for high complexity motor and motion control applications. With SmartFusion's integrated 32-bit ARM Cortex M3 MCU, programmable analog in-chip embedded nonvolatile memory and low power flash-based FPGA fabric, it addresses the challenges of motor control design.

 

Also at ESC Xilinx introduced an ARM Cortex processor-based platform that enables system architects and embedded software developers to apply a combination of serial and parallel processing with the horsepower required to drive tasks involving high-speed access to real-time inputs and complex digital signal processing neededd to meet their application-specific requirements.

 

It seems to me that the FPGA folks are making progress and in a year or tow they should make a real debate necessary.

 

What do you think?.

Filter Blog

By date: By tag: