From: John Stone (
Date: Mon Nov 24 2008 - 11:00:33 CST

  The existing Windows versions of VMD are standard 32-bit builds,
and thus they are limited to using 2GB of physical memory. It is
possible to compile applications to use extended memory addressing
modes that give access to about 3GB from a 32-bit app, but we haven't
bothered with this and have instead been working on building a 64-bit
version of VMD for Windows.

VMD itself does compile in 64-bit Windows builds, but there had been a
long period of problems with some of the libraries that VMD depends on.
I've heard from at least one individual that some of these problems
have been fixed, and so I will soon be trying to do a 64-bit build
of VMD for Windows again.

In the mean time, you'll have to work within the 2GB limitation,
either by using scripts like "bigdcd" to do analysis on large
trajectories, or by viewing subsets of your data by loading either
a selected range of trajectory frames, or by skipping frames,
e.g. loading only one out of every ten frames, for the purposes
of viewing your simulation.

I'm also working on adding support for working with out-of-core
datasets in VMD, but this is still in the early stages and won't
be done for some time yet.

  John Stone

On Mon, Nov 17, 2008 at 09:48:00AM -0800, wrote:
> Hi,
> I have a general question regarding the memory use of VMD here.
> In windows, is VMD capable of using the virtual memory or is it capable of
> storing large data file temporarily in harddisk? I am using VMD for the
> visualization of large configurations (200000 atoms in each frame, usually
> more than 500 frames in total) and I just have 4GB RAM on my desktop, so it
> would be a bad news for me if VMD can't use virtual memory.
> Thanks,
> Shuai

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:                 Phone: 217-244-3349
  WWW:      Fax: 217-244-6078