[mpich-discuss] Hydra binding questions

[Jain, Rohit]

> 1. We used taskset with Hydra. We are seeing about 20% slowdown with
> (2-1.4 Hydra + taskset) Vs (2-1.0.6 mpd + taskset) on same machine
> with same taskset binding. We were expecting Hydra to perform same or
> faster than mpd. Is that not true?

What kind of performance degradation are you seeing? If this is
application performance degradation, then mpd vs. hydra might have
little to nothing to do with it. This might be some totally separate
issue (e.g., something in the changes to MPICH2 from v1.0.6 to v1.4).

Also, you can use hydra's mpiexec from 1.4 to execute your application
compiled with mpich2-1.0.6. If you want to do a more head-to-head
comparison.

Rohit: I agree it may not have to do with mpd vs hydra. 
I am comparing same application with 2 different versions of MPICH. With using same binding methods, I wouldn't expect slowdown. There is 20% increase in runtime of our application with MPICH2-1.4, which is kind of surprising, as i was expecting similar or better runtime with newer MPICH.

I am also seeing 2-3% reduction in memory. I haven't traced it, but that is clearly due to newer MPICH2 version, and is good thing.