Details
- Deep Learning models work natively with NVIDIA hardware.
- Setonix at Pawsey contains a large number of AMD GPUs.
- Models can be made to work with AMD hardware available at Pawsey.
Completed
- Build AMD compatible container for AlphaFold2.
- Build AMD compatible container for ColabFold.
- Build AMD compatible container for Boltz-1.
- Build AMD compatible container for ESMFold.
- Build AMD compatible container for Boltz-2.
- Adapt containers for compatibility with nf-core/proteinfold.
- Test nfcore/proteinfold workflow at Pawsey.
- Contribute a ProteinFold-specific configuration to the nf-core Pawsey Setonix institutional profile.
Acknowledgements
This work was supported by Australian BioCommons and the Pawsey Supercomputing Research Centre. Australian BioCommons receives NCRIS funding through Bioplatforms Australia. Pawsey Supercomputing Research Centre receives funding from the Australian Government and the Government of Western Australia.