Re: [OMPI devel] OMPI vs Scali performance comparisons

[Eugene Loh]

Jeff Squyres (jsquyres) wrote: I still think that the pml fast path fixes would be good. As do I.  Again, I think one needs to go to the BTL sendi as soon as possible after entering the PML, which raised those thorny discussions about how exactly we must preserve current BTL ordering. To catch up to Scali, we need 2x.  The PML fast-path stuff I showed in Feb in San Jose produced 30%-3x, depending on the hardware (that is, depending on whether the bottleneck is memory performance or instruction-processing speed).  I don't know where this particular hardware sits in that range.