[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Discuss-gnuradio] 64 bit
From: |
Eric Blossom |
Subject: |
Re: [Discuss-gnuradio] 64 bit |
Date: |
Fri, 16 Dec 2005 14:40:41 -0800 |
User-agent: |
Mutt/1.5.6i |
On Fri, Dec 16, 2005 at 03:49:21PM -0500, cswiger wrote:
> Just curious - how much might gnuradio benefit from running
> on 64 bit processors and how much of a rewrite would it take,
> to take full advantage of, say, an Athlon 64 X2, seeing as you
> could theoretically chunk thru twice the data every bus clock?
It already runs on the X86-64, including a first cut at the SSE/3DNow
stuff. It's what I use everyday, and yes, it is substantially faster!
> I guess the Solaris build doesn't take advantage of the UltraSparc II's.
> A little benchmarking with a 300Mhz processor and a short graph:
>
> noise source --> throttle --> low pass filter --> null sink
>
> using 100Khz sample rate and 67 tap filter uses 80% of one cpu.
>
> --Chuck
The noise source is pretty expensive, particularly for GR_GAUSSIAN.
Also, you're using the generic fir filter kernel (straight C++). On
x86 and x86-64 we've got hand-coded assembler for taking advantage of
SSE and 3DNow SIMD instructions.
I don't know if there's a Solaris equivalent to "oprofile" (profiling
using the h/w performance counters), but the first step to making it
faster would be to determine where it's slow.
Eric