CBE Hyak Resources
For those interested in High-Performance Computing in the college, the CBE Dean’s Office and Office of Research have covered the cost of a single Hyak slice for college faculty and research staff to use to try out the system. Not only does this provide access to our node, but allows faculty/staff to run projects across all of the Hyak infrastructure when those resources are otherwise idle.
CBE Hyak Mail list For CBE Hyak users to share information/questions.
UW Hyak Resources
UW Hyak Mail list Used for service changes and interruptions. If you use Hyak, sign up.
UW Research Computing Resources
The Research Computing Club Facilitates access to and training for UW’s shared super computing cluster, Hyak, and cloud computing resources that are available to all undergraduate and graduate students. The RC Club provides access to these nodes and training students to use the effectively and responsibly.
The Hyak User Wiki has great information about using the Hyak system (we are on a KLONE node):
Storage
Hyak Storage General Information
Local: Each full node has 1.5TB or more of local NVME SSD disk storage. This is non-persistent storage and is cleared after a job ends. Data must be copied to and from local SSD before and after each job to utilize this.
Group: Each slice purchase includes 1 TB of storage space and a 1 million file count limit of shared group storage (i.e., gscratch) accessible from every node. Additional storage quota increases can be purchased for $10 per month for 1TB of additional space and 1 million additional file count limit. Additional “scrubbed” shared storage is available for short-term use, but will be automatically deleted if not accessed for several weeks.
Keep backup copies of anything you need in a separate location (which is a good plan in any case).
This was planned to be a shared resource for the college and for those wishing to check out the system. Access is by UW netID. To request access, send email to cbeit@uw.edu. The details on exactly how to share this resource is a learning curve for us, we are working through those.
Suggested reference links:
The Klone Data Commons is the cluster-wide, shared dataset storage located at /gscratch/data. The purpose of the Data Commons is to provide a single location for datasets being used by multiple groups, to avoid hosting the same dataset multiple times in separate group directories.