VMD-L Mailing List
From: John Stone (johns_at_ks.uiuc.edu)
Date: Wed Apr 06 2005 - 11:27:54 CDT
- Next message: jbaucom_at_unity.ncsu.edu: "capturing box dimensions from amber coordinate files"
- Previous message: John Stone: "Re: Displaying simulations and rotations"
- In reply to: Karsten Suhre: "Choice of a Linux machine for VMD and high-end computing"
- Next in thread: Sabuj Pattanayek: "Re: Choice of a Linux machine for VMD and high-end computing"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Karsten,
The answer to your question is remarkably simple actually:
1) For VMD in stereo on linux, the best choice I've tested recently
is NVidia Quadro. They far from perfect, but they can be made to behave :)
2) For the purposes of your BLAST jobs, you'll want a lot of RAM in this
machine, as it'll help cache disk I/O (random access I/O improvement).
Large RAM will also help VMD significantly.
3) For NAMD you want a machine with peak clock speed, great CPU cache
performance, etc.
Are you sure you want to do all three on one machine? I'd personally get
at least two or three different machines here, it'd probably cost less
than one really big 4-way SMP box anyway.... I have more commentary,
but I've gotta run for now.
John
On Wed, Apr 06, 2005 at 06:09:58PM +0200, Karsten Suhre wrote:
> Dear VMD-users and -developpers,
>
> given the ever recurring problems with hardware compatibilities, I would be
> interested in your suggestions concerning the selection of a new Linux
> machine, that what kind of graphics card to select / avoid, what kind of
> processor, motherboard, etc. would you recommend?
>
> I.e. do you know of a good configuration that can be used AT THE SAME TIME
> for high end computing (molecular dynamics with NAMD) and high throughput
> BLAST database searches (limitted by disk access, but very scalable) and for
> 3-D STEREO Visualization with VMD (incl. of course ray-tracing and the like).
>
> Ideally, I would be looking for a 4-processor board, but it seems there is
> no such with AGP or PCI-Express support (correct?).
>
> How about the processor - is a 64-bit Opteron a good choice? Or would you
> stick to 32 bit for compatibility reasons?
>
> Is NVIDIA still the "standard" for Stereo graphics under VMD?
>
> I would be very grateful for all comments, and I am sure others on this list
> will benefit from it as well,
>
> Kind regards,
>
> Karsten
-- NIH Resource for Macromolecular Modeling and Bioinformatics Beckman Institute for Advanced Science and Technology University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801 Email: johns_at_ks.uiuc.edu Phone: 217-244-3349 WWW: http://www.ks.uiuc.edu/~johns/ Fax: 217-244-6078
- Next message: jbaucom_at_unity.ncsu.edu: "capturing box dimensions from amber coordinate files"
- Previous message: John Stone: "Re: Displaying simulations and rotations"
- In reply to: Karsten Suhre: "Choice of a Linux machine for VMD and high-end computing"
- Next in thread: Sabuj Pattanayek: "Re: Choice of a Linux machine for VMD and high-end computing"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]