Supercomputers/Systems/Hera (LLNL)
USA · operational · hpc
Hera (LLNL)
Also known as Hera Appro Xtreme-X3
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
- 102.20 TFlop/s
- Rpeak
- 127.20 TFlop/s
- Rmax ÷ Rpeak
- 80.3%
- Cores
- 13,552
- 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.
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 Hera (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 | 102.20 TFlop/s | #66 of 136 · top 48% | #246 of 332 · bottom 27% |
| Rmax ÷ Rpeak | 80.3 % | #66 of 136 · top 48% | #104 of 331 · top 31% |
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 | - Appro Xtreme-X3 Server, quad Opteron quad-core | Appro HPENYSE
rolls up to Hewlett Packard Enterprise | - | Reported | top500.orgweb.archive.org |
| CPU | - AMD Opteron quad-core 2.3 GHz, 16 cores per node, 864 nodes | AMD | - | Reported | web.archive.orgtop500.org |
| Interconnect | Mellanox InfiniBand DDR Interconnect Fat tree · 20 Gb/s per 4x port · About 2 microseconds InfiniBand DDR (Mellanox) | Mellanox NVDANASDAQ
rolls up to NVIDIA | - | Reported | web.archive.orgtop500.org |
The read
Hera was a large capacity cluster shared by Livermore's M&IC and ASC programmes for small to moderate parallel jobs. LLNL's archived 2010 and 2011 Open Computing Facility pages list 864 nodes of four quad-core 2.3 GHz Opterons (16 cores, 32 GB each), 13,824 cores, 127.2 TFlop/s theoretical peak, InfiniBand DDR (Mellanox in the 2011 snapshot) and the CHAOS Linux distribution. TOP500 lists the Appro Xtreme-X3 at 13,552 cores, 102.2 TFlop/s Rmax and the same 127.2 TFlop/s Rpeak on nine lists, 32nd in June 2009 to 459th in June 2013; the 272-core difference to LLNL's figure is unexplained. It is unrelated to NOAA's later Hera (noaa-hera).
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 Hera (LLNL) added from its source pages
Nearby in scale
The machines closest to Hera (LLNL) by Rmax, smaller to larger.
- Olympus (PNNL) · USA, operational · 0.1 PFlop/s
- Red Storm · USA, decommissioned · 0.1 PFlop/s
- Eurora · Italy, operational · 0.1 PFlop/s
- Rostand · France, operational · 0.11 PFlop/s
- Chundoong · South Korea, operational · 0.11 PFlop/s
- Beacon (NICS) · USA, operational · 0.11 PFlop/s
Related systems
Same site
- Edge (LLNL) · USA, operational, 2010
- Atlas (LLNL) · USA, decommissioned, 2007
- Zin · USA, decommissioned, 2011
- Cab · USA, decommissioned, 2012
- Sequoia · USA, decommissioned, 2012
- Vulcan · USA, decommissioned, 2012
Same operator
- ASC Purple · USA, decommissioned, 2005
- BlueGene/L · USA, decommissioned, 2004
- Thunder · USA, decommissioned, 2004
- Jade (LLNL) · USA, operational, 2016
- MCR · USA, decommissioned, 2002
- Quartz (LLNL) · USA, decommissioned, 2016
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.