[Beowulf] CCL:Opteron or Nocona ? (fwd fromm.somers at chem.leidenuniv.nl)

Mon May 9 14:02:44 PDT 2005

I have just 1 question to Somers,

a) Did the AMD executables use SSE2?
b) Did you compare the dual core processors which have 2 processors for
every cpu?

I ask 'a', because default intel compiler doesn't allow AMD 
executables to use that. A small patch to the executable of it, removing
that check, should do the job, or just use another better compiler.

Additional it's not so difficult fine tuning SSE2 assembly code, if we
speak about just 1 routine eating all system time multiplying / calculating
invariant. A reasonable programmer needs perhaps 1 month time there and
entire department for many years to go can use that.

If i have a short look to the practical performance of some prime searching
software here at intel versus AMD (trying to win the $150k price as EFF
awards it, see: http://www.eff.org/awards/coop.html), the difference is
*so* huge in advantage of AMD. 

Of course itanium2 is faster in floating point, so if you can get those for
free from intel more or less it's an interesting cpu for such clusters,
otherwise it isn't.

There is many compilers that work for x86-64 nowadays at *nix OSes.

For code written by mediocre programmers for example real superior at
x86-64 behaving at several codes is the pathscale compiler. Might be a good
idea for you to use too, but try out all options. I remember someone toying
with it: "it didn't speed me up at all". Another one reports: "this and
that option combined speeds up my c++ code by 40%".

It is very interesting, to use polite wordings, to see when i hear someone
quote 40% speedup of pathscale over gcc 4.0.0, for a similar type program,
whereas a program doing similar things is faster in gcc.

Vincent

At 11:24 PM 5/9/2005 +0200, Eugen Leitl wrote:
>----- Forwarded message from mark somers <m.somers at chem.leidenuniv.nl> -----
>
>From: mark somers <m.somers at chem.leidenuniv.nl>
>Date: Mon, 9 May 2005 09:38:06 +0200
>To: Luigi Cavallo <cavallo at chemistry.unina.it>
>Cc: chemistry at ccl.net
>Subject: CCL:Opteron or Nocona ?
>User-Agent: KMail/1.6.2
>
>Dear Luigi Cavallo,
>
>we have recently purchased and setup a new Beowulf cluster here in Leiden, 
>dedicated to run ADF and other DFT codes (Dacapo and Vasp) on it. We have 
>found the Nocona's, in combination with the Intel compilers, to be the best 
>combination especially for ADF and ADF-BAND.
>
>Before we decided to go for the Nocona's, we did try with two test machines, 
>one being a dual-cpu AMD Opteron 246 (2.0 GHz) and the other being a
dual-cpu 
>Nocona (3.0 GHz). Tests showed that the AMD has, in general, a slightly 
>better cache hierarchy and a better memory scalability, but also that the 
>lack of compilers able to tune for the Opterons is severe. Of course we took 
>the differences in cpu clocks into account.
>
>As it turns out, after having contacted the SCM people in Amsterdam, ADF can 
>effectively use the SSE3 registers and cpu intruction set and that makes it 
>run fast on the Nocona's. This, together with the prices being roughly the 
>same, made us decide to go for the Nocona's. 
>
>BTW, maybe this is known to you already, but Intel offers you their OpenMP 
>capable compilers for free for academic use. 
>
>The machine is up and running now in 64 bit mode and doing fine ;-). 
>
>Maybe the information can help you in your decision.
>
>Best regards, 
>
>Mark Somers.
>
>> Hi,
>>
>> we are going to invest some money for a few computers, and we have to
>> make a decision between the AMD-Opterons and the Intel-Nocona. What's
>> better ? We are experienced with the Opterons, but we have no idea about
>> the Noconas...
>>
>> Major codes to run on them will be classical QM packages as ADF, G03, TM,
>> some AIMD as CPMD, and possibly some classical MD as gromacs.
>>
>> Thanks,
>> Luigi
>
>-- 
>Dr. M. F. Somers
>Theoretical Chemistry - Leiden Institute of Chemistry  - Leiden University
>Einsteinweg 55, P.B. 9502, 2300 RA Leiden, The Netherlands
>tel: +31715274437
>mail: m.somers at chem.leidenuniv.nl
>web:  http://rulgla.leidenuniv.nl/Researchers/Somers.htm
>room: HB322
>
>
>-= This is automatically added to each message by the mailing script =-
>To send e-mail to subscribers of CCL put the string CCL: on your Subject:
line
>and send your message to:  CHEMISTRY at ccl.net
>
>Send your subscription/unsubscription requests to: CHEMISTRY-REQUEST at ccl.net 
>HOME Page: http://www.ccl.net   | Jobs Page: http://www.ccl.net/jobs 
>
>If your is mail bouncing from ccl.net domain due to spam filters, please
>use the Web based form from CCL Home Page 
>-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>
>
>
>
>
>----- End forwarded message -----
>-- 
>Eugen* Leitl <a href="http://leitl.org">leitl</a>
>______________________________________________________________
>ICBM: 48.07078, 11.61144            http://www.leitl.org
>8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
>http://moleculardevices.org         http://nanomachines.net
>
>Attachment Converted: "g:\internet\eudora\attach\signature116.asc"
>_______________________________________________
>Beowulf mailing list, Beowulf at beowulf.org
>To change your subscription (digest mode or unsubscribe) visit
http://www.beowulf.org/mailman/listinfo/beowulf
>