[Beowulf] Re: Re: Home beowulf - NIC latencies

[Rossen Dimitrov]

> A last remark. I really think that the argument of using the same 
> swiss-army-knive MPI implementation such as ScaMPI or Intel MPI or even 
> MPI/Pro to infere interconnect characteristics is even worse that 
> looking at latency and bandwidth alone. These implementations are never 
> going to be designed to use all hardware efficiently, their design is 
> either historic (Scali used to provided software for SCI alone) or 
> politicaly motivated (Intel is using uDapl, hummm, wonder why), or both. 
> They are by-products of the MPI forum failure to make the Standard 
> practical (compatible ABI).
> 
> Patrick

Patrick, this is quite a broad statement. 4 years ago we had a paper 
arguing that MPI's written to support many different interconnects and 
messaging technologies through internal portability layers were probably 
sub-optimal for at least some of the interconnects. Most of the reasons 
are obvious. At the time we were dealing with Portals, LAPI, and GM. You 
can easily see why having an internal portability layer for these 
interfaces does not seem to easily match the semantics of either one of 
them. We probably did something in our design to reflect this.