Service Status

Description

The correct announcement can be seen below: WHEN IS IT HAPPENING? PACE's next Maintenance Period starts at 6:00AM Tuesday May 12th and is scheduled to end no later than 11:59PM on Friday May 15th. PACE will release each cluster (Phoenix, Firebird, and ICE) as soon as maintenance work and testing are completed. WHAT DO YOU NEED TO DO? As usual, jobs with resource requests that would be running during the Maintenance Period will be held until after the maintenance by the scheduler. During this Maintenance Period, access to all the PACE-managed computational and storage resources will be unavailable. This includes Phoenix, Firebird, and ICE. Please plan accordingly for the projected downtime. CEDAR storage will not be affected. WHAT IS HAPPENING? ITEMS REQUIRING USER ACTION/AWARENESS: [ICE, Phoenix] FY26 Refresh Node Trade-in: We are trading in old hardware to make space for the upcoming refresh of Phoenix and ICE – this will result in a temporary reduction in capacity, after which each system will have an increased number of cores and GPUs available. We are primarily trading in older generation or lower memory hardware. Recommendations for minimizing the impact of the trade-in process will be forthcoming. [ICE] All 4 NVIDIA V100-32GB GPU nodes will be removed. (All 11 NVIDIA V100-16GB GPU nodes will still be available, but since they rely on an older CUDA version, they will be removed as a part of the default GPU pool.) [ICE] NVIDIA Quadro RTX 6000s will remain available, but since they rely on an older CUDA version, they will be removed as part of the default GPU pool. [Phoenix] All AMD Milan CPU nodes and the pmem (Intel Optane) will be removed, as well as 36% of the cpu-small nodes. (Genoa AMD nodes will still be available via embers, and GNR nodes are an alternative resource. Information on requesting a GNR node is available here.) [all] Upgrade to Open OnDemand 4.1 There will be user interface improvements! All current applications will be supported on the new version. ITEMS NOT REQUIRING USER ACTION: [ICE] Upgrade ICE login nodes to new hardware [ICE] Reallocate one H100 GPU node from job scheduling to support upcoming AI services [Firebird] Migrate head nodes to new Granite Rapids nodes [Phoenix] Add new login node [Phoenix] Upgrade InfiniBand fabric subnet managers [Phoenix] Scratch storage lustre version upgrade [Phoenix] Introduce additional login node to production [all] Upgrade NVIDIA DGX firmware [all] Final integration and commissioning of secondary pump for research hall cooling loop [all] PACE Firewall Upgrades [all] Upgrade ethernet switch firmware [all] Slurm 25.11 upgrade [all] Upgrade storage system versions [all] Upgrade InfiniBand switch firmware [all] Upgrade kernel version and apply OS patches on all PACE supported servers WHY IS IT HAPPENING? Regular maintenance periods are necessary to reduce unplanned downtime and maintain a secure and stable system. WHO IS AFFECTED? All users across all PACE clusters. WHO SHOULD YOU CONTACT FOR QUESTIONS? Please contact PACE at pace-support@oit.gatech.edu with questions or concerns. You may read this message on our blog, available at https://pace.gatech.edu/posts/.

Components

Academic Services

Services

PACE

Schedule

May 12, 2026 6:00AM - May 15, 2026 6:00PM EDT