Computing systems
The Stony Brook Supercomputing Center supports a range of computational resources for the research community. The systems below are operated by SBSC and by Research Computing and Informatics, and each suits different work.
-
The flagship cluster of SBSC, Stony Brook's new Tier 1 supercomputing center. Funded by the National Science Foundation (award OAC-2505376) with more than 60,000 AMD cores for high-throughput science. Timeline, specifications, and launch-update sign-up on the Kitsune page.
-
The current generation of the SeaWulf cluster, adding GPU nodes and high bandwidth memory nodes to the general purpose pool.
- CPUs
- 48 AMD Milan, 94 compute Intel Sapphire Rapids w/ HBM
- GPU nodes
- 11, each with 4x A100 80GB GPUs and two Intel Skylake CPUs
-
The earlier generation of SeaWulf, still in service for general purpose simulation and modeling.
- CPUs
- 164 Intel Haswell, 64 Intel Skylake
- GPUs
- 28 Nvidia Tesla K80, 2 Tesla P100, 2 Tesla V100
- Storage
- 4 PB GPFS
-
A GPU cluster built for artificial intelligence and machine learning work, and the largest single expansion of campus AI computing capacity to date.
- GPUs
- 24 NVIDIA H200 NVL
- Peak
- 80 petaFLOPS FP8
-
A HIPAA compliant cluster for research involving protected health information.
- CPU cores
- 1,040
- GPU cores
- 104,448 CUDA, 13,056 Tensor
- Memory
- 8.4 TB RAM
- Storage
- 18 TB SSDs, 1 PB HDDs (NFS)
-
A National Science Foundation testbed for future supercomputing technology, and the first computer outside Japan built on the Fujitsu A64FX processor. Ookami left service at the end of 2025. The project record is kept for reference and citation.
- CPUs
- 176 Fujitsu A64FX each with 512 Gbyte SSD, 2 NVIDIA Grace Superchips, 2 Marvell ThunderX2, 1 Intel Skylake, 1 AMD Milan
- Storage
- 800 TB Lustre
- Funded by
- NSF OAC-1927880
