[MPICH] Does ssm make a difference?

[Benjamin Svetitsky]

Thanks, that's much improved.  The User's Guide led me to believe that 
ch3:ssm is to be preferred for multicore installations, but it gave no 
improvement at all over ch3:sock.

			Ben

Rajeev Thakur wrote:
> Try using ch3:nemesis instead. It should perform better.
> 
> Rajeev
> 
>> -----Original Message-----
>> From: owner-mpich-discuss at mcs.anl.gov 
>> [mailto:owner-mpich-discuss at mcs.anl.gov] On Behalf Of 
>> Benjamin Svetitsky
>> Sent: Thursday, January 10, 2008 1:54 AM
>> To: mpich-discuss at mcs.anl.gov
>> Subject: [MPICH] Does ssm make a difference?
>>
>> Dear mpich commmunity,
>>
>> I just installed MPICH2 on a cluster of Intel Core2 Quad 
>> cores.  First I installed with no options, and verified that 
>> it installed the ch3:sock device.  I ran benchmarks.  Then I 
>> reinstalled with ch3:ssm, and found that the benchmarks 
>> remained unchanged.  Isn't ssm supposed to enhance 
>> performance in multicore platforms?

-- 
Prof. Benjamin Svetitsky           Phone:  +972-3-640 8870
School of Physics and Astronomy    Fax:    +972-3-640 7932
Tel Aviv University                E-mail: bqs at julian.tau.ac.il
69978 Tel Aviv, Israel             WWW:    http://julian.tau.ac.il/~bqs