Supercomputers/Systems/HA-PACS
Japan · decommissioned · hpc
HA-PACS
Also known as Highly Accelerated Parallel Advanced system for Computational Sciences
Operated by University of Tsukuba at Tsukuba Center for Computational Sciences .
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
- 421.60 TFlop/s
- Rpeak
- 778.13 TFlop/s
- Rmax ÷ Rpeak
- 54.2%
- Cores
- 20,800
- Power
- 407.29 kW
- Per watt
- 1.0 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 | - Xtreme-X GreenBlade 8204 | Cray HPENYSE
rolls up to Hewlett Packard Enterprise | - | Reported | top500.org |
| Accelerator | - NVIDIA Tesla M2090 GPUs, four per node | NVIDIA | - | Reported | top500.orgccs.tsukuba.ac.jp |
| CPU | Intel Xeon E5-2670 CPUs x86-64 · 8C · 2.6 GHz · 32 nm · 115 W Two Intel Sandy Bridge-EP E5-2670 8C 2.6 GHz per node, 268 nodes | Intel | 268 reported | Reported | top500.orgccs.tsukuba.ac.jp |
| Interconnect | Mellanox InfiniBand QDR Interconnect Fat tree · 40 Gb/s per 4x port Dual-rail Infiniband QDR fat-tree | NVIDIA | - | Reported | top500.orgccs.tsukuba.ac.jp |
The read
The eighth-generation PACS/PAX machine at Tsukuba's Center for Computational Sciences. The base cluster has 268 nodes, each with two 8-core 2.6 GHz Sandy Bridge-EP Xeons and four NVIDIA M2090 GPUs, on dual-rail InfiniBand QDR fat-tree with 504 TB of Lustre storage, 802 TFlop/s peak as delivered in January 2012. TOP500 lists it 41st in June 2012 with 421.6 TFlop/s Rmax. A 64-node TCA extension with K20X GPUs was added in 2013 and is listed separately by TOP500.
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
Tsukuba Center for Computational Sciences
- Location
- Tsukuba, Japan
- Commissioned
- 1996
Change history
- 2026-10-02 Added HA-PACS, Tsukuba's GPU cluster
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 |
|---|---|---|
| ccs.tsukuba.ac.jp | names system, part and count | Mellanox InfiniBand QDR · counts: 268 |
| 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.