[Beowulf] Athlon64 / Opteron test

[Lombard, David N]

From: Mikhail Kuzminsky
> According to Joe Griffin
[deletia]
> >
> > MSC.Nastran Hardware comparison:
> >
>
http://www.mscsoftware.com/support/prod_support/nastran/performance/v04_
sn
> gl.cfm
> >
>   This page contains very interesting tables w/description of hardware
> used, but at first look I found only the data about OSes, not about
> compilers/run time libraries used. The (relative bad) data for IBM
> e325/Opteron 2 Ghz
> looks "nontrivial"; I beleive some interptretation of "why?" will be
> helpful.
> May be some applications used are relative cache-friendly and have
working
> set
> placing in large Itanium 2 cache?

For the record, on these single-processor MSC.Nastran jobs, FP
performance, memory bandwidth, and I/O bandwidth are most significant;
memory latency is irrelevant.  MSC.Nastran is decidedly not
cache-friendly; having said that, numerical kernels generally use rank-n
updates, n>1, with the actual rank size being selected via
processor-specific tuning.

> May be it depends from compiler and Math library used ? BTW, for LGQDF
> test:
> I/O is relative small (compare pls elapsed and CPU times which are
very
> close);

LGQDF does a TB of I/O, as shown in the table at the top.

> but Windows time for Dell P4/3.2 Ghz (4480 sec) is much more worse
than
> for Linux on the same hardware (3713 sec). IMHO, in this case they
>  must be very close in the case of using same comlilers&libraries
>  (I don't like Windows, but this result is too bad for this OS :-))

Note: I left MSC.Software last year; I ceased running the Porting
organization (i.e., having direct contact with how MSC.Nastran was
built) in '00.

Same machine differences like the Dell Linux/Windows results are more
often OS related.  See especially XXAFST, a statics job dominated by a
single math kernel, with relatively low I/0 (yes, 77GB is "low" for
MSC.Nastran), where the Windows and Linux CPU times are very close.
Contrast that with the three jobs with the greatest Windows/Linux time
difference, they do the largest amounts of I/O.

-- 
David N. Lombard
 
My comments represent my opinions, not those of Intel Corporation.