From: John Stone (johns_at_ks.uiuc.edu)
Date: Wed Oct 17 2007 - 17:28:37 CDT

Hi,
  Ah, you're using MacOS X, the current MacOS X versions of VMD are all 32-bit
due to fact that Apple's OpenGL is (up to 10.4 anyway) only 32-bit.
This limits the amount of memory you can allocate to less than 4GB in
a single application, and since the kernel takes some of the virtual address
space, it may even be down around 2GB depending on the system. So, in
short, you're going to have to load this huge DX file on another system
where you can run a 64-bit copy of VMD. Just loading the DX file you've
got will break 2GB, and by the time you create a graphical representation
like an Isosurface, you're going to need even more. There's basically no
way around it, you'll need a 64-bit machine, or you'll need to downsample
your DX file to a lower resolution version in order to work with it on
that system.

Cheers,
  John Stone
  vmd_at_ks.uiuc.edu

On Wed, Oct 17, 2007 at 06:09:54PM -0400, Corenflos, Steven Charles wrote:
> I had the researcher run this at a lab on campus that has a 64-bit Mac OS X 10.4.something on which they were willing to install VMD for me. I don't know if he used the debug flag but here was the output:
>
> Last login: Wed Oct 17 15:27:01 on ttyp1
> Welcome to Darwin!
> 129-79-20-162:~ avl$ /Volumes/VMD-1.8.6alpha/VMD\ 1.8.6.app/Contents/MacOS/startup.command; exit
> Info) VMD for MACOSXX86, version 1.8.6 (April 6, 2007)
> Info) http://www.ks.uiuc.edu/Research/vmd/
> Info) Email questions and bug reports to vmd_at_ks.uiuc.edu
> Info) Please include this reference in published work using VMD:
> Info) Humphrey, W., Dalke, A. and Schulten, K., `VMD - Visual
> Info) Molecular Dynamics', J. Molec. Graphics 1996, 14.1, 33-38.
> Info) -------------------------------------------------------------
> Info) Multithreading available, 4 CPUs detected.
> Info) OpenGL renderer: NVIDIA GeForce 7300 GT OpenGL Engine
> Info) Features: STENCIL MDE CVA MTX NPOT PP PS GLSL(OVF)
> Info) Full GLSL rendering mode is available.
> Info) Textures: 2-D (4096x4096), 3-D (4096x4096x4096), Multitexture (4)
> vmd > RCNE SendEventToEventTarget (suom Moved ) failed, -50
> NavGetReply failed, -128
> RCNE SendEventToEventTarget (suom Moved ) failed, -50
> Info) Analyzing Volume...
> Info) Grid size: 313x625x625 (1865 MB)
> Info) Total voxels: 122265625
> Info) Min: -102.365997 Max: 343.471405 Range: 445.837402
> Info) Computing volume gradient map for smooth shading
> VMD(251,0xa000d000) malloc: *** vm_allocate(size=1467191296) failed (error code=3)
> VMD(251,0xa000d000) malloc: *** error: can't allocate region
> VMD(251,0xa000d000) malloc: *** set a breakpoint in szone_error to debug
> terminate called after throwing an instance of 'std::bad_alloc'
> what(): St9bad_alloc
> /Volumes/VMD-1.8.6alpha/VMD 1.8.6.app/Contents/MacOS/startup.command: line 7: 251 Abort trap "$p/../Resources/VMD.app/Contents/MacOS/VMD" $*
> logout
>
>
> He says it seems to fill up to about 1.6GB of memory, grinds away, then crashes.
> Does this help or shall I attempt it again with the debug flag myself?
>
>
> -----Original Message-----
> From: owner-vmd-l_at_ks.uiuc.edu on behalf of Axel Kohlmeyer
> Sent: Mon 10/15/2007 4:56 PM
> To: Corenflos, Steven Charles
> Cc: vmd-l_at_ks.uiuc.edu
> Subject: Re: vmd-l: Problems Opening Large .dx File
>
>
> On Mon, 15 Oct 2007, Corenflos, Steven Charles wrote:
>
> SC> First of all I apologize if I don't convey the necessary
> SC> information. I'm doing technical support for a chemistry lab which
> SC> has brought me to this mailing list.
>
> hi steve,
>
> don't worry about this. there are probably quite a few people
> around that understand your situation very well.
>
> SC> I open VMD and go to File->New Molecule. I have two molecules I'm
> SC> working with: one is a 177 MB .dx file, and the other is a 3.5 GB
> SC> .dx file. The 177 MB file (we can call it A), opens just fine.
> SC> Attempting to open the 3.5 GB file (which we'll call B) causes top
> SC> to show 100% CPU usage on one of the CPUs, and variable memory usage
> SC> usually in the range of 15-17%. This continues for several minutes
> SC> until the program spontaneously terminates.
>
> in order to get more specific information about when and where
> VMD crashes, you could start VMD in the debugger using the -debug flag.
>
>
> SC> I'm running this on a dual-processor, dual-core Xeon 3.06 GHz server with 6 GB of RAM, running Gentoo Linux.
>
> are you running a 32-bit or a 64-bit OS and VMD version? with the 32-bit
> you may be simply running out of memory...
>
> SC> In case it's relevent I also want to point out that it's only using
> SC> 100% of one core; top shows only 25% total CPU usage. Also I'm
>
> VMD is only in parts multi-threaded and definitely not when reading
> data files.
>
> SC> running this on one of our servers because the same file caused an
> SC> out-of-memory error on the workstation that was originally
> SC> processing it, which had only 1 GB of RAM. I realize it
> SC> theoretically should be able to allocate any additional memory it
> SC> needs from swap, but since one file opens and the other doesn't (and
>
> again, this can be 32-bitness issue. in 32-bit more x86 machines can
> handle up to 64 GB in total, but are still limited to 32-bit in the
> available address space per process.
>
> SC> since I wasn't 100% certain vmd wasn't allocating memory directly
> SC> instead of through system calls) I decided to try it on our server.
>
> SC> Please let me know what additional information I can provide to help figure this problem out.
>
> just let us know, if the above is helpful and see if you can get
> the additional info.
>
> cheers,
> axel.
>
> SC>
> SC> Thanks,
> SC> -Steve
> SC>
>
> --
> =======================================================================
> Axel Kohlmeyer akohlmey_at_cmm.chem.upenn.edu http://www.cmm.upenn.edu
> Center for Molecular Modeling -- University of Pennsylvania
> Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
> tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425
> =======================================================================
> If you make something idiot-proof, the universe creates a better idiot.
>
>

-- 
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