Processor contention(?) and network bandwidth on AMD

[Martin Siegert]

On Mon, Apr 29, 2002 at 03:40:30PM -0400, Mark Hahn wrote:
<snip>
> NOTE TO ALL BEOWULF USERS: seriously consider running 2.4.18
> or better, and *definitely* try out gcc 3.1 ASAP.  
> 
> these updates have changed my life </evangelical>

I can only confirm that 2.4.18 improves "life" dramatically.
Before that certain MPI jobs (using LAM) would just hang.

Now with respect to gcc-3.x:
It has been pointed out that gcc-3.0 is extremely bad for Athlons
and (only) bad for PIIIs:
http://math-atlas.sourceforge.net/errata.html#gcc3.0
My recent exercise in compiling atlas only confirms this.

Have these issues been resolved in gcc-3.1?

Martin

========================================================================
Martin Siegert
Academic Computing Services                        phone: (604) 291-4691
Simon Fraser University                            fax:   (604) 291-4242
Burnaby, British Columbia                          email: siegert at sfu.ca
Canada  V5A 1S6
========================================================================