Supercomputers/Systems/Ansel (LLNL)
USA · operational · hpc
Ansel (LLNL)
Also known as Ansel; RZAnsel (probable)
Operated by Lawrence Livermore National Laboratory at Livermore Computing Center .
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
- 1.29 PFlop/s
- Rpeak
- 1.69 PFlop/s
- Rmax ÷ Rpeak
- 76.3%
- Cores
- 19,656
- 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 | - IBM Power System AC922 | IBM | - | Reported | top500.orghpc.llnl.gov |
| Accelerator | NVIDIA Tesla V100 Accelerators CUDA (Volta GV100) · 5,120C · 1.53 GHz · 12 nm · 300 W NVIDIA Tesla V100 (Volta), four per compute node, 216 GPUs per LLNL's RZAnsel page | NVIDIA | 216 reported | Reported | hpc.llnl.govtop500.org |
| CPU | IBM POWER9 22C CPUs POWER · 22C · 3.07 GHz · 14 nm · 190 W IBM POWER9 22C 3.1GHz (LLNL page: 44 cores per node, 2,376 cores in total) | IBM | - | Reported | top500.orghpc.llnl.gov |
| Interconnect | Mellanox InfiniBand EDR Interconnect Fat tree · 100 Gb/s per 4x port · About 0.9 microseconds Dual-rail Mellanox EDR InfiniBand (LLNL page: IB EDR) | Mellanox NVDANASDAQ
rolls up to NVIDIA | - | Reported | top500.orghpc.llnl.gov |
| Operating system | Red Hat Enterprise Linux Operating systems Linux distribution · Open source, commercial support subscription Red Hat Enterprise Linux | Red Hat | - | Reported | top500.orghpc.llnl.gov |
The read
Small Sierra-generation IBM system at Lawrence Livermore National Laboratory. TOP500 lists it as Ansel: IBM Power System AC922, POWER9 22-core 3.1 GHz, dual-rail Mellanox EDR InfiniBand, NVIDIA Tesla V100, Red Hat Enterprise Linux, installed 2018, 19,656 cores, Rmax 1.29 and Rpeak 1.69 PFlop/s, with no power or HPCG figure, listed from November 2018 (247th) to June 2020 (471st, its last list). LLNL's platform page for RZAnsel, an ATS-2 / CORAL-1 class system commissioned in 2018 in the open (RZ) zone, gives 62 nodes (one login), 2,376 POWER9 cores, 216 V100 GPUs (four per compute node), InfiniBand EDR and 1.570 PFlop/s peak, and marks it decommissioned without a date. TOP500's 19,656 cores equal 2,376 CPU cores plus 17,280 GPU cores (80 per GPU across 216 GPUs), so the entry is very probably RZAnsel listed without the zone prefix, but neither page says so and the peaks differ (1.69 versus 1.570). Status is therefore held at the last TOP500 list rather than LLNL's decommissioned label for RZAnsel. The larger siblings are Sierra and Lassen.
Timeline
Public list appearances
Pointers only: rank and edition, linking to the canonical entry. We do not reproduce list tables. See sourcing policy.
Site & facility
Livermore Computing Center
- Location
- Livermore, CA, USA
- Cooling
- Direct liquid cooling (warm water)
- Commissioned
- 1996
Change history
- 2026-10-18 Added Ansel (LLNL)
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.