autodock-gpu 1.5.2 does not work on Cardinal
autodock-gpu/1.5.2 does not work on Cardinal. We suggest using autodock-gpu/1.6 instead
Software
Classroom Jupyter and RStudio Apps
If you have classroom projects at OSC that used the Jupyter or RStudio apps through class.osc.edu before June 30, 2025 and plan to reuse your classroom environment, the PI will need to complete the following one‑time update:
Wine32 Segmentation Fault
Users have reported that running Wine32 (32-bit Wine) results in a Segmentation fault (core dumped). The issue is linked to how cgroups (control groups) are handled or enabled on specific compute nodes or environments.
starccm does not use vis node GPU
Issue:
When trying to run starccm on a vis node in order to use GPU for visualization (not for GPU calculation), the typical way to do this would be to load the virtualgl module and use vglrun
module load virtualgl/3.1.1
module load starccm/19.06.009
vglrun starccm+
However, starccm does not pick up libraries from $LD_LIBRARY_PATH by default, so it does not pick up the necessary graphics libraries
vLLM versions prior to 0.14.1 are deprecated
vLLM versions prior to 0.14.1 are deprecated due to security issue CVE-2026-22778 which can allow remote code execution. Clients are advised to use versions 0.14.1+ with module load vllm/0.14.1
Apptainer apt GPG Error Caused by Private /tmp
Duplicate of:
https://www.osc.edu/resources/technical_support/known_issues/apptainer_c...
On any compute node, users may encounter the following error when installing packages with apt in Apptainer:
GPU/VIS nodes for various OOD apps are broken
After the most recent downtime we discovered that various OOD apps relying on the "virtualgl" module on Cardinal were broken. We have since updated and pinned the latest virtualgl version for each broken app
Apptainer container builds fail on compute nodes due to /tmp namespace behavior
Building Apptainer containers on compute nodes may fail during apt update or other package operations. The failure is caused by Apptainer’s build environment not being fully isolated from the host when private /tmp namespaces are in use.
Schrodinger 2023.2 does not launch
Schrodinger 2023.2 does not launch through OnDemand or OnDemand desktop, failing with "fatal error"
If you do not need 2023.2 specifically, or you are able to transition to 2024 or later versions, please do so. We highly recommend this option over the following workarounds
Schrodinger can be launched on vis nodes, through X11Forwarding, and through batch jobs. If trying to run Schrodinger 2023.2, please use a vis node, use X11Forwarding, or do not use the GUI and run through a batch job
AlphaFold 3 GPU Out-of-Memory Error During Inference
When you run AlphaFold 3, you may encounter a GPU out-of-memory (OOM) failures during model execution. The job terminated with errors similar to:
