Hi,
This should be fine; see
https://docs.nvidia.com/deploy/cuda-compatibility/index.html
for the details.
Best regards,
Takanori Nakane
On 4/20/24 21:03, Victor Banerjee wrote:
> Hey Takanori, thanks! I'll get on that. I've noticed something weird
> though. When I type 'nvidia-smi' it shows CUDA 12.4, but my workstation
> only has CUDA 11.8 installed (/usr/local). When I try to install relion
> during the 'make' step, it says it found CUDA 11.8. Could that be
> causing the problem?
>
> Thanks,
> Victor
>
>
> On Fri, Apr 19, 2024 at 4:02 PM Takanori Nakane
> <[log in to unmask] <mailto:[log in to unmask]>>
> wrote:
>
> Hi,
>
> Why are you using such an old version of RELION?
>
> The best way forward is to compile the latest version of RELION
> with the right version of CUDA SDK.
> Make sure RELION is NOT linked to packages (CUDA, MPI runtime etc)
> provided by other software suites (e.g. CCPEM, conda, CryoSPARC)
> by checking PATH and LD_LIBRARY_PATH. Use `which` and `ldd`.
>
> Best regards,
>
> Takanori Nakane
>
> On 4/20/24 04:36, Victor Banerjee wrote:
> > Hi,
> > Recently, we've encountered an issue with our workstation when
> attempting to run GPU-optimized jobs in Relion3.1. Specifically,
> we're encountering the following error:
> > "ERROR: out of memory in
> /usr/local/relion-3.1-source/src/acc/cuda/custom_allocator.cuh at
> line 436 (error-code 2)
> > in: /usr/local/relion-3.1-source/src/acc/cuda/cuda_settings.h,
> line 67
> > ERROR: A GPU-function failed to execute.
> >
> > If this occurred at the start of a run, you might have GPUs which
> are incompatible with either the data or your installation of
> Relion. If you:
> >
> > -> INSTALLED RELION YOURSELF: if you e.g. specified
> -DCUDA_ARCH=50 and are trying to run on a compute 3.5 GPU
> (-DCUDA_ARCH=3.5), this may happen."
> >
> > A few weeks ago, we installed Cryosparc, but we're uncertain if
> this installation is related to the issue, as Relion was functioning
> properly subsequent to the installation. The workstation is equipped
> with four GPUs (RTX A6000). Even when attempting to utilize the
> tutorial data, we're encountering the same error. We've attempted to
> resolve the issue by reinstalling Relion on the workstation,
> unfortunately without success.
> >
> > I'm reaching out to inquire if anyone has experienced a similar
> issue and if they might have any insights or recommendations to
> resolve it. Any assistance or guidance you could provide would be
> greatly appreciated.
> > Thank you for your attention to this matter.
> >
> > Best regards,
> >
> >
> ########################################################################
> >
> > To unsubscribe from the CCPEM list, click the following link:
> > https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCPEM&A=1
> <https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCPEM&A=1>
> >
> > This message was issued to members of www.jiscmail.ac.uk/CCPEM
> <http://www.jiscmail.ac.uk/CCPEM>, a mailing list hosted by
> www.jiscmail.ac.uk <http://www.jiscmail.ac.uk>, terms & conditions
> are available at https://www.jiscmail.ac.uk/policyandsecurity/
> <https://www.jiscmail.ac.uk/policyandsecurity/>
>
########################################################################
To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCPEM&A=1
This message was issued to members of www.jiscmail.ac.uk/CCPEM, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
|