[Beowulf] many cores and ib

Gilad Shainer Shainer at mellanox.com
Tue May 20 09:58:59 PDT 2008

Patrick Geoffray wrote:
> > My apologizes. I meant the MPI includes an option to 
> collect several 
> > MPI messages into one network message. For applications cases, 
> > sometimes it helps with performance and sometimes it does not. OSU 
> > have shown both cases, and every user can decide what works 
> best for him.
> Please, point me to a document/paper/report that shows that 
> message coalescing improves real applications. Even from 
> Panda, that's fine.

I am glad you approve Panda. Many thanks. They have publish a nice paper
on the usage of message coalescing, and I have already send it to
several folks who asked for it. Since you are very smart, it will be
easy for you to find it by yourself. Some homework will only do good for

> > As long as you use the same code and the same benchmark 
> (and the same 
> > platform), you can use any benchmark to compare between different 
> > network devices. You can claim that if you want to count 
> the network 
> > messages and not the MPI messages, using this way to do it 
> will give 
> > you the wrong number, but if you want to compare between 2 
> > interconnects, this is valid.


> If you want to compare, do it with Open-MPI then, it does not 
> yet implement message coalescing. Same benchmark, same 
> platform, same MPI. 
> That should work, no ?

I believe that message coalescing is in Open MPI 1.3. Didn't we talked
on doing some homework? .... 

> > This becomes personal now... :-) you don't need to be angry all the 
> > time. Bad for your health. I could have replay back 
> mentioning all the
> People often mistake anger for passion. 

Naaa... It is anger, and it is really bad for you. 

> I have to say, your 
> case (marketing abusing technical info) is a very special 
> treat. 

So abusing technical info is what you have done in your paper on RDMA?
Just want to make sure I understand what you meant. And speaking on
that, it was very funny to see Renato respond back with "A Tutorial of
the RDMA Model" article especially for you. I am sure we mention
homework already, right? ... 

> I handle it very much like chocolate: a little bit at 
> a time, only once a week. 
> Don't worry for my health, though I appreciate your concern.
> > FUD and misleading information you provide wherever you go, 
> and we saw
> > couple of examples just few weeks ago, but I wont.   
> I guess you are referring to the technical panel at the IDC 
> forum in Norfolk and my correction about the 
> (in)compatibility between iWARP and IB. If so, please explain 
> how the last-byte-written-last semantic in iWARP and IB are 
> the same. 

Did someone said anything on interoperability between iWARP and
InfiniBand? I did not hear that. The only mentioning was that Open
Fabrics provide SW solution for both IB and iWARP. Try to listen before
responding. It helps sometimes.

> Please, give me a sweet one, with something 
> chewy/gooey in the middle. I just love those :-)

Be careful so it will not stick to your teeth, as it might keep your
mouth shut

> Patrick

More information about the Beowulf mailing list