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

[Vincent Diepeveen]

At 11:07 14-2-2005 -0800, Greg Lindahl wrote:
>On Mon, Feb 14, 2005 at 06:47:15PM +0300, Mikhail Kuzminsky wrote:
>
>> Let me ask some stupid's question: which MPI implementations allow
>> really
>>  
>> a) to overlap MPI_Isend w/computations
>> and/or 
>> b) to perform a set of subsequent MPI_Isend calls faster than "the 
>> same" set of MPI_Send calls ?
>> 
>> I say only about sending of large messages.
>
>For large messages, everyone does (b) at least partly right. (a) is
>pretty rare. It's difficult to get (a) right without hurting short
>message performance. One of the commercial MPIs, at first release, had
>very slow short message performance because they thought getting (a)
>right was more important. They've improved their short message
>performance since, but I still haven't seen any real application
>benchmarks that show benefit from their approach.

Perhaps no one who needed fast latency bought those NICs in the first place.

A huge number of jobs that the 1024 processor SGI of dutch government used
to handle is 4-8 processors. Simply because latency matters.

>-- greg
>_______________________________________________
>Beowulf mailing list, Beowulf at beowulf.org
>To change your subscription (digest mode or unsubscribe) visit
http://www.beowulf.org/mailman/listinfo/beowulf
>
>