Supercomputers/Systems/CP-PACS
Japan · decommissioned · hpc
CP-PACS
Also known as Hitachi SR2201/2048
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
- 368.2 GFlop/s
- Rpeak
- 614.0 GFlop/s
- Rmax ÷ Rpeak
- 60.0%
- Cores
- 2,048
- 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 | - Hitachi SR2201 | Hitachi | - | Verified | museum.ipsj.or.jpnetlib.org |
| CPU | Hitachi SR2201 processing unit CPUs PA-RISC 1.1 · 1C · 0.15 GHz Hitachi SR2201 PU | Hitachi | 2,048 reported | Verified | museum.ipsj.or.jpnetlib.org |
| Interconnect | Hitachi 3D hyper-crossbar Interconnect 3D hyper-crossbar · 300 MB/s point to point 3D hyper-crossbar | Hitachi | - | Verified | netlib.orgmuseum.ipsj.or.jp |
| Storage | Distributed RAID 5 disk array Storage Direct-attached, no separate metadata server · Hundreds of GB, era-dependent 529 GB distributed disks in RAID 5 configuration | Hitachi | - | Verified | museum.ipsj.or.jpnetlib.org |
| Cooling | Air cooling (CRAC/CRAH) Cooling Room-level air handling · Practical ceiling around 15-20 kW per rack multichip module now with fins for air-cooling | Hitachi | - | Reported | arxiv.org |
| Operating system | Hitachi HI-UX/MPP Operating systems Unix (Mach 3.0 microkernel), Hitachi-customised · Proprietary HI-UX/MPP | Hitachi | - | Reported | museum.ipsj.or.jptop500.org |
The read
A university machine at the top of the list, which has not happened since. Tsukuba had been building the PACS/PAX series since the 1970s and co-designed CP-PACS with Hitachi, so the SR2201 that Hitachi then sold commercially carries technology developed for a physics department. The processor is Hitachi’s own implementation of HP’s PA-RISC with pseudo-vector hardware bolted on, which is a licensing arrangement the parts table has to represent as design nationality separate from instruction-set origin. It ran for nine years, retiring in September 2005.
Timeline
- 1991 Announced source
- 1996-10 First operational source
- 1996-11 List debut at #1 (November 1996) source
- 2005-09 Decommissioned source
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
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 |
|---|---|---|
| museum.ipsj.or.jp | names system, part and count | Hitachi SR2201 processing unit, Hitachi 3D hyper-crossbar, Hitachi HI-UX/MPP, Distributed RAID 5 disk array · counts: 2048 |
| netlib.org | names system, part and count | Hitachi SR2201 processing unit, Hitachi 3D hyper-crossbar, Hitachi HI-UX/MPP, Distributed RAID 5 disk array · counts: 2048 |
Checked 2026-10-09 by pnpm verify. Re-run it and the table changes with the web.
Further reading
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.
Operating a supercomputer
Acceptance testing, why "installed" and "in production" are six months apart, and the machine lifecycle.