Extended Summer Downtime (July/Aug 2025)

[koomie]

After two years of production and numerous GPU upgrades, the AI & HPC Cluster is physically moving to a new data center. The new location will allow for improved power/cooling infrastructure to accommodate future upgrades and networking improvements but will incur an extended system downtime while the entire cluster is deconstructed, packed, shipped, and reassembled in the new data center.

Specific downtime dates and durations are not finalized just yet, but we expect the downtime to begin around July 15th, 2025 and last for 2-4 weeks. During the transition, the entire cluster (including login node) will remain unavailable. Prior to the downtime, users are encouraged to download any files/results needed to their local system as the cluster file
systems will also remain unavailable.