Michael> Sounds convincing. So the improvement is probably due to
Michael> better send/receive parallelism we are getting in mthca.
I won't have time to try it until next week but it should be easy to
test the QP lock split without the IPoIB CQ split and see how that compares.
- R.
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
