Supercomputers/Systems/Lynx (LLNL)
USA · operational · hpc
Lynx (LLNL)
Also known as LLNL/NNSA CTS-2 Lynx
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 1581 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
- 5.65 PFlop/s
- Rpeak
- 6.45 PFlop/s
- Rmax ÷ Rpeak
- 87.6%
- Cores
- 100,800
- Power
- 986.45 kW
- Per watt
- 5.7 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.
How it compares
Ranked against the other machines in this dataset that report the same figure: those that first came into service within two years of Lynx (LLNL), and those in USA. Computed from our tables; a metric is ranked only when at least eight peers report it.
| Metric | Value | Against its era | Against its country |
|---|---|---|---|
| Rmax | 5.65 PFlop/s | #119 of 158 · bottom 26% | #75 of 332 · top 22% |
| Energy efficiency | 5.7 GF/W | #77 of 81 · bottom 6% | #41 of 162 · top 25% |
| Rmax ÷ Rpeak | 87.6 % | #28 of 155 · top 18% | #36 of 331 · top 11% |
| Power draw | 986 kW | #33 of 85 · top 38% | #60 of 171 · top 35% |
Every metric is ranked largest-first, so a high power-draw rank means a big power draw, not a good one. Rmax ÷ Rpeak uses HPL-measured results only. Rankings move as machines are added; they describe this dataset, not the world.
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 | - Dell PowerEdge C6620 servers (CTS-2) | Dell Technologies | - | Reported | top500.orgstoragenewsletter.com |
| CPU | - Intel Xeon Platinum 8480+ 56C 2 GHz | Intel | - | Reported | top500.org |
| Interconnect | - Cornelis Networks CN5000 400G Omni-Path fabric, 952 nodes | Cornelis Networks | - | Reported | storagenewsletter.comtop500.org |
| Operating system | Tri-Lab Operating System Stack Operating systems Linux distribution, RHEL-derived, HPC-tuned · Open source, LLNL-maintained Linux/TOSS (Tri-Laboratory Operating System Stack) | Lawrence Livermore National Laboratory | - | Reported | top500.org |
The read
Livermore's newest Commodity Technology Systems-2 (CTS-2) cluster: a 952-node system of Dell PowerEdge servers with Intel Xeon processors and the Cornelis CN5000 400G Omni-Path fabric, which Cornelis announced as deployed at ISC 2026 (June 2026), saying it adds production capacity for NNSA's Advanced Simulation and Computing program and wider national security work. TOP500 (June 2026, 277th; the list is full for our table so no list row is added) lists a Dell PowerEdge C6620 with Xeon Platinum 8480+ 56C 2 GHz, Cornelis CN5000 and TOSS, installation year 2026, 100,800 cores, Rmax 5.65 and Rpeak 6.45 PFlop/s and 986.45 kW. 100,800 cores is 900 dual-socket 56-core nodes, fewer than the 952 nodes Cornelis cites, so the benchmark run appears to have used part of the machine (our inference, not stated by either source). Sibling CTS-2 systems at the lab are Dane and Bengal (llnl-dane, llnl-bengal).
Site & facility
Livermore Computing Center
- Location
- Livermore, CA, USA
- Cooling
- Direct liquid cooling (warm water)
- Commissioned
- 1996
Change history
- 2026-10-17 Added Lynx (LLNL)
Nearby in scale
The machines closest to Lynx (LLNL) by Rmax, smaller to larger.
- Hazel Hen · Germany, decommissioned · 5.64 PFlop/s
- Sirius (Tsukuba) · Japan, operational · 5.63 PFlop/s
- COBRA · Germany, decommissioned · 5.61 PFlop/s
- JMA Cray XC50 · Japan, operational · 5.73 PFlop/s
- Sawtooth · USA, operational · 5.78 PFlop/s
- Theta · USA, decommissioned · 5.88 PFlop/s
Related systems
Same site
- Neo · USA, operational, 2026
- Matrix · USA, operational, 2025
- Tuolumne · USA, operational, 2025
- El Capitan · USA, operational, 2024
- rzAdams · USA, operational, 2024
- Bengal · USA, operational, 2023
Same operator
Source check
We fetched this system's own citations and recorded whether each page actually mentions it. This is a corroboration signal, not a fact check, and it is published so you can see how well the sourcing holds up rather than take it on trust.
| Cited source | Result | Found on the page |
|---|---|---|
| storagenewsletter.com | does not name this system | - |
| top500.org | names system only | - |
Checked 2026-10-18 by pnpm verify. Re-run it and the table changes with the web.
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.