Re: SMT on IBM Power5+

try to search here for "smt":
http://www-03.ibm.com/support/techdocs/atsmastr.nsf/Web/Techdocs

There is no special document for POWER5+. But I don't think there is a
difference to POWER5 with SMT.

Cheers,

Andy


On Wed, 04 Apr 2007 20:05:11 -0500, "Michael E. Thomadakis"
<miket@xxxxxxxxxxx> wrote:

Jan-Frode Myklebust wrote:
On 2007-04-04, Michael E. Thomadakis <miket@xxxxxxxxxxx> wrote:
More specifically, I would expect that SMT ON would benefit workload
mixes with high thread concurrency but with low contention on common
thread resources (L1/L2/L3 and functional units). For instance, I would
expect a cluster with a large number of distributed (and multithreaded)
processes running to benefit from SMT ON: twice as many h/w threads can
be active at a time, obviating the need for costly thread-switch.

I did some benchmarking on a P5+ cluster last fall, and was
quite surprised that SMT=ON (and 1 task per virtuall cpu vs.
1 task per physical cpu in SMT=OFF) was giving best performance
on all the benchmarks we did. Even for 16 node (~250 task)
mpi-jobs.


-jf

Jan thanks for your reply. We have a 40 node p5-575 cluster with 32GB
DRAM/node. Nodes are attached to 2 planes of HPS. I was contending that
SMT ON should benefit the cluster since all of the cluster control
processes (RSCT, GPFS, HPS assistance, etc.) are heavily multi-threaded
and by not being compute-intensive, they should benefit from SMT ON.

Are there any public benchmarks that I can run to demonstrate the
benefits of SMT ?

thanks again,
Michael


--
Andreas Beckmann
Andreas.Beckmann@xxxxxxxxxxx
http://www.muenster.de/~andy
.

Relevant Pages

  • Re: SMT on IBM Power5+
    ... Andreas Beckmann wrote: ... There is no special document for POWER5+. ... For instance, I would expect a cluster with a large number of distributed processes running to benefit from SMT ON: twice as many h/w threads can be active at a time, obviating the need for costly thread-switch. ...
    (comp.unix.aix)
  • RE: DMX Content Query
    ... You could call the stored procedure used by the cluster viewers. ... either says options are not allowed in sub-select, or smt like that... ...
    (microsoft.public.sqlserver.datamining)
  • Re: Intel might revive Hyperthreading with Nehalem
    ... caused by the trace cache limitation you mention. ... So you're claiming that the P4 SMT is not *real* because it does round ... half-assed or less than 'true' multithreading. ... I guess according to Yousuf this means that the POWER5 doesn't have ...
    (comp.sys.ibm.pc.hardware.chips)
  • Re: Simultaneous multithreading
    ... > I am new to SMT. ... SMT is what IBM call a technology on POWER5. ... cpu is packaged on a chip. ... a cpu between two threads and get about 30% extra performance (on ...
    (comp.unix.aix)
Loading