Summary
Two resources for free computing at UW-Madison offered by the Center for High Throughput Computing (CHTC):
– BadgerCompute: badgercompute.wisc.edu/ for Jupyter-based interactive notebooks in R, Python, or Julia computer programming languages.
– High Throughput Computing for large compute projects. Free account by filling-in a request account form.
Free computing @ UW
There was a time when there was no free computing available at UW-Madison, which makes the current options even more valuable. Here are 2 valuable resources, meant for different kind of computing offered by the Center for High Throughput Computing (CHTC.)
BadgerCompute
Free to UW-Madison. Web site: badgercompute.wisc.edu/
No installation necessary to run code in R, Python, or Julia computer programming languages. The web-based system uses the Jupyter-based interactive notebooks which allows computation and visualization.
A jupyter notebook can take advantage of Markdown easy to understand structured text for beautiful formatting and documenting along with the code. (See my previous blogs “Do Yourself a Favor…” to learn Markdown.)
BadgerCompute is great for academic or research settings (Policies) for:
- testing scripts using Python, R, or Julia
- exploring and visualizing datasets
- writing or following tutorials written in Jupyter Notebooks
Getting Started
It is necessary to enroll every year and you can only log in to BadgerCompute if you have completed the Canvas Course as described in the “Get Started” page. There are 3 steps:
Step 1: Enroll in Canvas Course
Step 2: Complete modules in course
Step 3: Account created within 24 h after course completion.
High Throughput Computing
For larger computing requirements, the CHTC offers and advanced, large Linux-based cluster. As of January 2024, there are approximately 40,000 CPU slots and 80+ GPU slots available in the CHTC execute pool.
UW-Madison personel can access a free account by filling-in a request account form. Students are also allowed with the approval of their supervising Professor.
Jobs are run in non-interactive fashion using the HTCondor (job) scheduler, which is also used within the Biochemistry Compute Cluster (BCC.) Biochemistry users can request a BCC account and train on using HTCondor with my quickstart tutorial, my previous blog Guide to Submitting HTCondor jobs, and more resources on my mini course *nix Tutorials – HTCondor.
Image credits: CHTC and BadgerCompute web sites.