Supercomputers/Systems/Gordon
USA · decommissioned · hpc
Gordon
Operated by SDSC at San Diego Supercomputer Center .
Phase 1 seed dataset, compiled by hand. These rows were built from public operator, laboratory and vendor sources. A mechanical second-reader pass has since fetched every cited source: 1092 of 1153 systems have a readable citation that names them, and 40 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
- 218.10 TFlop/s
- Rpeak
- 262.25 TFlop/s
- Rmax ÷ Rpeak
- 83.2%
- Cores
- 12,608
- Power
- 252.16 kW
- Per watt
- 0.9 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 | - Appro Xtreme-X GreenBlade GB512X (TOP500 now lists the manufacturer as Cray/HPE, reflecting Cray's 2012 acquisition of Appro's HPC business) | Cray HPENYSE
rolls up to Hewlett Packard Enterprise | - | Reported | top500.orgsdsc.edu |
| CPU | Intel Xeon E5-2670 CPUs x86-64 · 8C · 2.6 GHz · 32 nm · 115 W Xeon E5 (Sandy Bridge-EP) 8C 2.6GHz, 12,608 cores in the debut configuration | Intel | - | Verified | top500.orgsdsc.edu |
| Interconnect | Mellanox InfiniBand QDR Interconnect Fat tree · 40 Gb/s per 4x port InfiniBand QDR | Mellanox NVDANASDAQ
rolls up to NVIDIA | - | Reported | top500.org |
| Storage | - Over 300 terabytes of Intel SSD 710 Series flash across 64 I/O nodes, rated at 35 million IOPS | Intel | - | Reported | sdsc.edusdsc.edu |
The read
SDSC's flash-memory-heavy NSF system, among the first supercomputers built around massive solid-state storage rather than spinning disk, delivering 35 million IOPS from over 300 terabytes of Intel SSDs across 64 I/O nodes. Built by Appro as the Xtreme-X GreenBlade GB512X (TOP500 now attributes it to 'Cray/HPE' after Cray bought Appro's HPC business in 2012), it debuted 48th on the November 2011 TOP500 in a partially populated test configuration of 12,608 Xeon E5-2670 cores at 218.10 TFlop/s, the figures recorded here; SDSC's own announcement independently states the same '218 teraflops ... 12,608 cores', so the tier is verified. Full production from January 2012 reached 16,384 cores and over 280 TFlop/s per SDSC's own later release, close to but not identical with TOP500's subsequent entries for the same system id (16,160 cores, 285.80 TFlop/s); that fuller configuration never out-ranked the debut entry, so it is recorded here in prose rather than as the headline figures. Gordon's five-year NSF allocation ended 31 March 2017, after which the Simons Foundation's Flatiron Institute took over at least 90 percent of its capacity for a further two years; no source states when it was finally powered down, so no decommissioned date is recorded.
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
Change history
- 2026-09-23 Added with a rank pointer on its debut edition; later full-production numbers recorded only in prose
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 |
|---|---|---|
| sdsc.edu | names system and part | Intel Xeon E5-2670 |
| top500.org | names system and part | Intel Xeon E5-2670, Mellanox InfiniBand QDR |
Checked 2026-10-09 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.