VMD-L Mailing List

From: John Stone (johns_at_ks.uiuc.edu)
Date: Fri Apr 22 2005 - 11:06:09 CDT

Kos,
  There are two issues involved in loading multi-gigabyte trajectories
on a given computer:
  1) 64-bit file I/O operations (can you read a file bigger than 2GB)
  2) 64-bit memory addressing (if you load the _whole_ dcd into memory)

If you're using a recent version of Linux, you can _read_ DCD files
larger than 2GB, regardless of what processor you have. This doesn't
mean that you can load the whole file into memory, but it means you can
use the 'bigdcd' script to perform analysis on it one-frame-at-a-time just
like Charmm.

And in case you haven't read my previous email, there _IS_ a way to run
VMD in text mode:
  vmd -dispdev text

That runs VMD without a graphics display, so you can use it for analysis.
If you know you're not going to issue any rendering commands during
the analysis session (even with no graphics window VMD can still
render images using Tachyon etc...) you will want to issue the command
"display update off" which save VMD a little bit of extra work.

The text mode way of running VMD won't use much less memory than the
full graphics mode. Most of the memory consumed is for storing the
trajectory data. If you load the data, the memory will be used, simple
as that. If you need to save memory, then use the 'bigdcd' script to
process your trajectories a frame at a time.

  John

On Fri, Apr 22, 2005 at 11:43:11AM -0400, Kos wrote:
> Hi,
>
> about DCD_splitters I think that we should try CHARMM
> or is there any other free suggestion please?
> Bogdan please explain what do you mean about CHARMM problems by 'I/O bandwidth and passes' (examples HW)?
>
> I think that in the past John Stone emphasized that for larger than 2GB files one needs 64bit environment(both hardware and OS?)
> Our LINUX clusters are AMD dual K7 ATHLON and INTEL dual XEON (130nm,32bit)
> so they don't support 64bit technology.
> We bought recently an HP laptop ZD8000,17" with the new Intel chipset; 64bit technology (650 HT).Also Windows 64bit are available but even with Windows XP(32) could I use a version of VMD-win that supports 64bit machines?
>
> Why there is not a VMD option to do only math-calculations eg RMS,g(r), torsion analysis (no graphics required)and then don't load graphically trajectories, pdb files
> saving a lot of memory and enchance speed please?
> Let's say only command line VMD-env.
>
> Kos
>
>
>
> Bogdan Costescu <Bogdan.Costescu_at_iwr.uni-heidelberg.de> wrote:
>
> >
> >On Thu, 21 Apr 2005, Kos wrote:
> >
> >> Is CHARMM an alternative solution to these problems and can treat
> >> large files faster because doesn't need graphics?
> >
> >A bit simplified view is that CHARMM might not need much memory for
> >analysis because it can load the trajectory frame by frame, so at any
> >given time only one frame is in memory. The disadvantage is that it
> >makes the analysis limited by the I/O bandwidth of the computer (to
> >local disk or NFS server) when you want to do several passes through
> >the whole trajectory; if you load the whole trajectory in RAM, it's
> >much faster for the second or subsequent passes.
> >
>
>
>
> __________________________________________________________________
> Switch to Netscape Internet Service.
> As low as $9.95 a month -- Sign up today at http://isp.netscape.com/register
>
> Netscape. Just the Net You Need.
>
> New! Netscape Toolbar for Internet Explorer
> Search from anywhere on the Web and block those annoying pop-ups.
> Download now at http://channels.netscape.com/ns/search/install.jsp 

-- 
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