Supercomputers/Systems/Intel Paragon XP/S 140
USA · decommissioned · hpc
Intel Paragon XP/S 140
Also known as XP/S 140
Operated by Sandia National Laboratories at Sandia National Laboratories, Albuquerque .
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
- 143.4 GFlop/s
- Rpeak
- -
- Rmax ÷ Rpeak
- -
- Cores
- 3,680
- 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 | - Intel Paragon XP/S 140 | Intel | - | Verified | netlib.orgosti.gov |
| CPU | Intel i860/XP CPUs i860 RISC · 1C · 0.05 GHz · 1000 nm Intel i860/XP | Intel | 3,680 reported | Verified | netlib.orgosti.gov |
| Interconnect | Intel Paragon 2D mesh Interconnect 2D mesh with torus wraparound · 200 MB/s per link Paragon XP/S 2-D mesh | Intel | - | Verified | netlib.orgosti.gov |
| Storage | Paragon I/O node RAID storage Storage Direct-attached RAID behind dedicated I/O nodes, diskless compute nodes reliable cost-effective mass storage via arrays of commodity 3.5-in disks protected by hardware RAID | Intel | - | Reported | archive.org |
| Cooling | Air cooling (CRAC/CRAH) Cooling Room-level air handling · Practical ceiling around 15-20 kW per rack Air-cooled operation | Intel | - | Reported | archive.org |
| Operating system | SUNMOS Operating systems Custom lightweight kernel, no Unix layer · Open source, government-authored SUNMOS | Sandia National Laboratories | - | Verified | osti.govnetlib.org |
The read
The first massively parallel machine to take the top of the list outright, and the moment the vector era stopped being the default answer. It is also the clearest early example of an operator supplying part of its own stack: Intel’s OSF/1 would not scale, so Sandia and the University of New Mexico wrote SUNMOS, a kernel under 256KB, and ran the machine on it. SUNMOS became PUMA, PUMA became Cougar, and Cougar ran ASCI Red on this same site three years later, which is a supply-chain lineage that no spec string records. No public source gives a decommissioning date, so the field is empty rather than guessed.
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
Sandia National Laboratories, Albuquerque
- Location
- Albuquerque, NM, USA
- Cooling
- Chilled water
- Commissioned
- 1990
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 |
|---|---|---|
| netlib.org | names system and part | Intel i860/XP, Intel Paragon 2D mesh, SUNMOS, Paragon I/O node RAID storage |
| osti.gov | could not be fetched | - |
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.