Supercomputers/Systems/BNL Institutional Cluster Gen2
USA · operational · mixed
BNL Institutional Cluster Gen2
Also known as IC gen2; IC2
Operated by Brookhaven National Laboratory at Brookhaven National Laboratory (Upton) .
Researched with AI assistance, checked by scripts. These rows were researched by AI agents reading public operator, laboratory and vendor sources, appended by scripts and checked by automated validation; no subject-matter expert has reviewed them (how this site is made). A mechanical second-reader pass has since fetched every cited source: 1350 of 1540 systems have a readable citation that names them, and 64 are genuinely weakly sourced. Every claim carries its source and a confidence tier. Treat anything below verified as a lead, not a citation.
Measured performance
- Rmax
- -
- Rpeak
- -
- Rmax ÷ Rpeak
- -
- Cores
- 2,544
- Power
- -
- Per watt
- - GF/W
Figures are as last publicly reported for the configuration described below, not a live measurement. Where a system was upgraded in place, the post-upgrade configuration is shown and the earlier one appears in the timeline.
What this machine is made of
One row per supplier relationship. “Supplier at build” is the company that shipped the part at the time; where that company has since been acquired, the parent it rolls up to today is shown beside it. That distinction is what makes ticker-level aggregation possible across a thirty-year dataset.
| Role | Part | Supplier at build | Quantity | Confidence | Source |
|---|---|---|---|---|---|
| Integrator | - Supermicro SYS-610C-TR CPU nodes, SYS-220GQ-TNAR+ and SYS-120GQ-TNRT GPU nodes | Supermicro | 53 reported | Reported | docs.sdcc.bnl.gov |
| Accelerator | NVIDIA A100 SXM4 80GB Accelerators 4x NVIDIA A100-SXM4-80GB in each of 12 GPU nodes | NVIDIA | 48 reported | Reported | docs.sdcc.bnl.gov |
| Accelerator | - 2xA100 PCIe (80 GB and 40 GB) debug nodes amperehost01 and amperehost02 | NVIDIA | 4 reported | Reported | docs.sdcc.bnl.gov |
| CPU | - Intel Xeon Gold 6336Y 2.40GHz 24-core, dual socket per node | Intel | 106 reported | Reported | docs.sdcc.bnl.gov |
| Interconnect | NVIDIA InfiniBand NDR (Quantum-2) Interconnect Fat tree or dragonfly+ · 400 Gb/s per port · Under 0.6 microseconds InfiniBand NDR200 connectivity on all nodes | NVIDIA | - | Reported | docs.sdcc.bnl.gov |
| Storage | - 1 PB GPFS distributed storage | IBM | - | Reported | docs.sdcc.bnl.gov |
The read
Slurm cluster at Brookhaven's Scientific Data and Computing Center that followed Institutional Cluster gen1; the 2024 USQCD all-hands site report says the KNL, IC and SKY clusters were retired on 1 October 2023, leaving three HPC clusters at the centre. Documentation lists 39 CPU nodes (Supermicro SYS-610C-TR, dual Xeon Gold 6336Y, 512 GB), 12 GPU nodes (SYS-220GQ-TNAR+, dual Xeon Gold 6336Y, 1 TB, four A100-SXM4-80GB each) and two debug nodes with two A100 PCIe cards each (one 80 GB, one 40 GB), all on NDR200 InfiniBand with 1 PB of GPFS and 1.9 TB local disk per node, so 53 nodes of 48 cores (2,544 CPU cores) and 52 A100 GPUs. Partitions are reserved: cfn (35 CPU and 8 GPU nodes), csi (4 GPU nodes), lqcd (4 CPU nodes) and debug (2 GPU nodes). The same report notes lattice QCD kept only 4 CPU nodes on IC gen2 after the retirements and was procuring a new cluster. No HPL result is published.
Timeline
No dated events recorded.
Site & facility
Change history
- 2026-10-18 Added BNL Institutional Cluster Gen2 (Brookhaven National Laboratory)
Further reading
Operating a supercomputer
Acceptance testing, why "installed" and "in production" are six months apart, and the machine lifecycle.
Inside a node
Latency-optimised versus throughput-optimised processors, NUMA, and the end of the host-to-device copy.
The interconnect
Topologies, why latency and tail behaviour matter more than bandwidth, and how the fabric market consolidated.