Re: [ccp4bb] xeon, 64 bit etc

- Protein crystallography

Main steps:

- Protein purification
- Crystallisation

Special:

- Programs for crystallography
- X-ray detectors

Basic tutorials:

   - Chemistry
   - Protein
   - Peptide
   - Amino Acids

Xtal community:

- CCP4BB

CCP4bb navigation

CCP4bb <-- 1999 <-- November 1999 <-- 30 November 1999

Previous message:
Subject: Re: Query on program for creating random crystallization screen
From: Patrick Shaw Stewart Patrick {- at -} DOUGLAS {- dot -} CO {- dot -} UK
Date: 2008-09-12

Next message:
Subject: Re: interface energetics
From: Steven Darnell sdarnell {- at -} BIOCHEM {- dot -} WISC {- dot -} EDU
Date: 2008-09-12

Subject: Re: xeon, 64 bit etc
From: Serge Cohen s {- dot -} cohen {- at -} NKI {- dot -} NL
Date: 2008-09-12

Hi Jan;

Indeed you consider buying more machines with less cores per CPU:
In multi-core, the memory bandwidth is limited per CPU, (the current
specific case of intel is even worth, there a strong limit on the
shared CPU bandwidth).

If the process you plan to be running are memory bandwidth bound, you
will be better off with multiple machines. The main difficulty is to
know if you are in such a situation, the only realistic way to know is
to experiment!

Now that I have answered the non asked question, here's my opinion on
your question : unless you have very large problem, using 32bits
executable will be safer. Still you can have a large memory >4GB and a
64bits system (the system will use all the memory, but each process
will have acces to max 4GB -indeed likely less, but at least 2GB-).

Hope this helps.

Serge.

Le 12 sept. 08 ¨¤ 18:18, Jan Abendroth a
¨¦crit :

> Hi all,
> we are looking into a multicore Xeon number cruncher with loads of
> memory. Are there any known issues we should consider regarding
> large memory (>4GB), 32bit vs. 64bit, and ccp4 or other
> crystallographic programs running on these systems? Only thinking
> linux at this point...
>
> Thanks a lot for any input!
> Jan
>
> --
> Jan Abendroth
> deCODE biostructures
> Seattle / Bainbridge Island, WA, USA
> work: JAbendroth_at_decode.com
> home: Jan.Abendroth_at_gmail.com

CCP4bb navigation

CCP4bb <-- 1999 <-- November 1999 <-- 30 November 1999

Previous message:
Subject: Re: Query on program for creating random crystallization screen
From: Patrick Shaw Stewart Patrick {- at -} DOUGLAS {- dot -} CO {- dot -} UK
Date: 2008-09-12

Next message:
Subject: Re: interface energetics
From: Steven Darnell sdarnell {- at -} BIOCHEM {- dot -} WISC {- dot -} EDU
Date: 2008-09-12