Hi Unity Users! We hope you had a wonderful holiday and a happy new year! The Unity Digest is back with some updated documentation and key announcements.
UPDATED DOCS ✏️
- Updated our documentation on conda environments to reflect the change in module name
ANNOUNCEMENTS 📢
💡 Unity profile is now available for Nextflow nf-core pipelines
An institutional Unity profile is now available for Nextflow nf-core pipelines (docs, config)!
To use it, add the -profile unity
option to any nf-core pipeline. This allows for each process to be submitted as a slurm job and use apptainer
for dependency management.
⚙️ New GPU and CPU additions
We’ve made some exciting updates to accommodate new CPU and GPU additions as part of the URI expansion:
- NVIDIA H100
- One node with 4x Nvidia H100 (80GB VRAM each) with 1TB of CPU RAM is now in the general
gpu
queue - An additional H100 node is available in the
gpu-preempt
queue - To access these, use
h100
with--gpus
or--constraint
- One node with 4x Nvidia H100 (80GB VRAM each) with 1TB of CPU RAM is now in the general
- NVIDIA L40s
- Eight additional L40s have been added in
gpu-preempt
(note: these are 32-cores)
- Eight additional L40s have been added in
- CPUs
- Four 64-core, 250GB CPU nodes were added to
cpu
- Twelve such nodes added to
cpu-preempt
- Four 64-core, 250GB CPU nodes were added to
🏆 Get involved with Harmony!
You might remember that before the holidays, the Unity team received an NSF grant to add Harmony to the Unity Research Compiting Platform. You can now read the full article on the Harmony award here.
If you’re interested in getting involved — as a student worker, beta tester, or simply to receive updates when deployment is complete — please fill out this form: https://forms.office.com/r/phehJTxHww.
❄️ We hope everyone had a wonderful holiday!
The Unity Team has returned to regular help hours.