Re: [Discuss-gnuradio] UC professor says SW radio a bad idea

[Eric Blossom]

On Mon, Feb 04, 2002 at 11:35:04PM -0800, Steve Schear wrote:
> Party's over for low-voltage CMOS, academics say
> By Stephan Ohr, EE Times
> Feb 4, 2002 (9:15 AM)
> 
> URL: http://www.eetimes.com/story/OEG20020204S0027
> 
> Other speakers in the evening session, concentrating on the effects of 
> low-voltage on digital circuits, offered conciliatory messages about the 
> consequences. Bob Brodersen, the University of California professor who 
> leads the Berkeley Wireless Research Center, pointed out that dedicated 
> function architectures using large amounts of parallelism offered the 
> highest efficiency - in terms of millions of operations per second (Mops) 
> per milliwatt - per unit area of silicon.

OK.  This argues that for devices where power consumption is
everything, highly parallel stuff is more efficient.  No surprise.
Not everything has this constraint.

> His thesis was based on an 
> examination of the processors presented at ISSCC over the past 20 years. 
> The software-intensive general-purpose processors with high clock rates 
> faired the worst in terms of Mops accomplished per milliwatt, he said. DSPs 
> with parallel math operations show a more efficient use of current and 
> voltage, or more Mops per mW. 

Of course, the most efficient portable DSP designs are also hand coded
in assembler.  This is because current compiler technology can't deal
with the highly irregular instruction sets, multiple mode bits,
specialized functional units, etc.

> But the most efficient semiconductor devices 
> - those demonstrating four orders of magnitude efficiency improvement - 
> were dedicated processors for MPEG-2 and 802.11 decoding. Such reasoning 
> rules against a general-purpose processor - and software-intensive 
> operations - for portable systems. "The software radio is a really bad 
> idea," Brodersen concluded.

We shall see...

Eric