Greg Lindahl wrote:
little computation. InfiniPath gets a speedup on lots of codes that
you wouldn't predict given the raw latency and bandwidth; how else
would you explain it?
There are a tons of variables. The one I keep thinking about is PIO
sending for larger message size than usual. If the data is in cache
(reasonable assumption for send side), it can remove a lot of load from
the memory bus compared to DMA. If your code is memory bandwidth bounded
(aren't they all on multi-core ?), then you have a speedup.
Ok, my turn to bite :-) What is a negative "g" ?
It means that the interconnect is ready to send a 2nd message before
the 1st one is on the wire. Think pipelining. Or you could ask
That's a warping of the (old and getting older) logp model :-) g cannot
be negative, the best it could be is null, which means the messages will
be send on the wire with no bubble between them. You cannot use a
negative g to express a NIC overhead lower than host, because a negative
g would compensate o for a single core, and it's not true.
Patrick
_______________________________________________
Beowulf mailing list, Beowulf@beowulf.org
To change your subscription (digest mode or unsubscribe) visit
http://www.beowulf.org/mailman/listinfo/beowulf
