[Beowulf] Slection from processor choices; Requesting Giudence

[Joachim Worringen]

laytonjb at charter.net wrote:

> Sure. This is for one specific CFD code. Let's assum that MPICH 1 is the basline.
> MPICH2 was twice as fast (half the run time). LAM was 30% faster than MPICH2.
> Scali MPI Connect was 20% faster than LAM. Altogether, Scali MPI Connect
> was twice as fast as MPICH1.

Hmmm...

T_mpich2= 0.5*T_mpich1
T_lam   = 0.7*T_mpich2
T_scali = 0.8*T_lam

I'd say this means:

T_scali = 0.28*T_mpich1

which is nearly four times as fast!?

  Joachim

-- 
Joachim Worringen - NEC C&C research lab St.Augustin
fon +49-2241-9252.20 - fax .99 - http://www.ccrl-nece.de