Protein folding (AlphaFold) at SCALE using a notebook and the cloud

Brendan Bouffler
Date : April 20, 2022
Categories : AWS ParallelCluster ,Amazon NICE DCV ,Life Sciences
Tags : cpus ,covid 19 ,dcv ,ec2 ,gpus ,hcls ,hpc ,high performance computing ,lustre ,parallel cluster ,schedulers ,storage ,alphafold ,protein folding ,supercomputing ,virtualization ,vizualization ,techshorts

Protein folding simulations can consume huge CPU time on supercomputers. AlphaFold changed that by applying ML techniques to the task. Then the problem became ‘how do I run AlphaFold easily?’.

The Healthcare AI/ML team at AWS figured that out, built a really accessible solution and open-sourced it to GitHub, which is where you can grab it today.

Brian Loyal and Ujjwal Ratan from the AI/ML specialist team in our global HCLS org joined us to help understand how this fits, why it works, and described some of the awesome science it supports.

They blogged about it here: https://hpc.news/curie264 The CloudFormation stack is here: https://github.com/aws-samples/aws-batch-architecture-for-alphafold

If you have ideas for technical topics you’d like to see us cover in a future show, let us know by finding us on Twitter (@TechHpc) and DM’ing us with your idea.