Alex Granovsky
gran@classic.chem.msu.su
>I've i Linux Suse Cluster:
>with following details :
>Core i7 / Linux Firefly version running under Linux.
> Running on Intel CPU: Brand ID 0, Family 6, Model 26, Stepping 5
> CPU Brand String : Intel(R) Xeon(R) CPU X5560 @ 2.80GHz
> CPU Features : CMOV, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, HTT, MWAIT, EM64T
> Data cache size : L1 32 KB, L2 256 KB, L3 8192 KB
> max # of cores/package : 8
> max # of threads/package : 16
> max cache sharing level : 16
> actual # of cores/package : 4
> actual # of threads/package : 8
> actual # of threads/core : 2
> Operating System successfully passed SSE support test.
>
>
> PARALLEL VERSION (MPICH) RUNNING IN SERIAL MODE USING SINGLE PROCESS
>48 Nodes ,each node 2 cpus ,and each cpu 4 cores
>what should i use as command line and input for best performance (MP2)
>(i tried lots of options for best performance but it seems the performance is
>still slow and it doesnt calculate on more than 1 node)
The most important is that you need to run Firefly in parallel.
The details how to do this depends on MPI you are using and
queue system (if any). You may need to consult with the system
administrator of this cluster.
You also need to use just:
$p2p p2p=.t. dlb=.t. $end
as the mxbuf settings you have used are good only for shared
memory version of P2P protocol, not for the tcp/ip based one.
regards,
Alex Granovsky