The subMIT computing facility is a login pool that is designed to provide access to the basic research computing resources of the physics department and beyond.
This one day workshop will provide an overview and updates on the status and plans for the system and project, as well as topical presentations on a range of use cases.
Additionally, there will be tutorial sessions. Bring your laptop to run the examples in real-time or just sit back and listen to the accompanying instructional talk.
More information on the subMIT project as well as user documentation is available at https://submit.mit.edu
The workshop will take place in the Kolker Room and a Zoom connection will be available at https://mit.zoom.us/j/96743699673?pwd=b3h2Q3c3cVQwYW12blhMUG5SWXZCZz09
Refreshments will be provided during coffee breaks!
Please note: the details of the schedule of individual talks may change as we work to accommodate new requests for contributions.
Computational Resources
Documentation: Users Guide, GitHub Examples
Support: Help Desk & Chat Bot
In this session we will outline key changes in the migration to the new version of Linux as well as how to access the older version via containers. We are also setting aside a portion of time to answer user questions about the upgrade, so please bring any questions you have, general or specific.
This session will show how to use the batch schedulers/resource-managers SLURM & HTCondor to manage your computational tasks, distributing them across the shared resources. Depending on your workflow, this may provide an easy way to run your calculations in parallel, shortening time-to-result and eliminating some manual tasks.
This tutorial will give an overview of using pytorch lightning for building and training neural networks. A simple problem in inference: measuring the parameters of a line will be presented. The tutorial will also introduce distributed training with pytorch lightning