VMD-L Mailing List
From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Tue May 06 2014 - 12:58:13 CDT
On Tue, May 6, 2014 at 12:08 PM, Erik Nordgren <nordgren_at_sas.upenn.edu>wrote:
> Hi folks,
> I have what must be a very common question, although I tried searching the
> list archives and somehow didn't come up with anything very recent &
> relevant, so figured I'd post.
> Basically, I'm just wondering if folks who are accustomed to purchasing
> new hardware regularly could comment with thoughts on the "optimum" choice
> (in terms of power vs. cost) of a GPU to put in a desktop workstation
> today, for smooth visualization of VMD structures with, say, 100-200 K
> atoms. (I assume that the "sweet spot" for choosing a GPU is a moving
> target, with the ever-improving capabilities of cards, which is why posts
> on this subject from over a year ago are probably not very relevant
> anymore.) I should add that I'm not in the market for an entire brand-new
> workstation, but rather considering just upgrading the GPU in the linux box
> I already have (a Dell Precision T3500, few years old already), which at
> the moment has an NVIDA Quadro NVS 295.
i have a 5 year old desktop and am not missing anything because it has a
*lot* of RAM. particularly for visualization and analysis RAM is the most
important asset. i recently popped in a (donated) GeForce TITAN to replace
a (donated) GeForce GTX 480, but for almost all VMD work i am doing
(occasionally with a lot of atoms), i could not tell much of a difference
in performance. until you enable GLSL rendermode (default is basic OpenGL),
you will see even less difference, as more work is done on the CPU side
with basic OpenGL rendering. almost any mid-level geforce and beyond should
give you a significant improvement over the (rather minimal) nvs card.
> As a related question, is it true that the only GPU manufacturer worth
> seriously considering for VMD is NVIDIA (due to the CUDA optimizations)?
currently only very few and very compute intense tasks have CUDA support.
for the most part the preference for nvidia is due to their rather complete
OpenGL support in the drivers. other vendor drivers were often flaky or not
at all usable for OpenGL. over the last few years, however, the situation
for open source drivers has improved, but not to the level of performance
of the closed source vendor drivers. if you intend to also use the GPU for
CUDA or OpenCL tasks, you cannot use the open source drivers (currently).
> Many thanks in advance for any & all suggestions!
> C. Erik Nordgren, Ph.D.
> Department of Chemistry
> University of Pennsylvania
-- Dr. Axel Kohlmeyer akohlmey_at_gmail.com http://goo.gl/1wk0 College of Science & Technology, Temple University, Philadelphia PA, USA International Centre for Theoretical Physics, Trieste. Italy.