Supercomputers/Systems/Pegasus
Brazil · operational · hpc
Pegasus
Also known as Pegaso
Operated by Petrobras .
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
- 19.07 PFlop/s
- Rpeak
- -
- Rmax ÷ Rpeak
- -
- Cores
- 233,856
- Power
- 1.03 MW
- Per watt
- 18.5 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 | - Atos (listed by TOP500 as Bull) with Supermicro A+ Server 4124GO-NART+ nodes | Atos | - | Reported | nextplatform.comtop500.org |
| Integrator | - Supermicro A+ Server 4124GO-NART+ GPU servers | Supermicro | - | Reported | top500.orgblog.nvidia.com.br |
| Accelerator | - 2,016 NVIDIA A100 80GB GPUs | NVIDIA | 2,016 reported | Verified | blog.nvidia.com.brtop500.org |
| CPU | AMD EPYC 7513 (Milan) CPUs x86-64 · 32C · 2.6 GHz · 7 nm · 200 W AMD EPYC 7513 32C 2.6GHz | AMD | - | Reported | top500.org |
| Interconnect | NVIDIA InfiniBand HDR Interconnect Fat tree or dragonfly+ · 200 Gb/s per port · About 0.6 microseconds Infiniband HDR | NVIDIA | - | Reported | top500.org |
| Operating system | CentOS Operating systems Linux distribution, RHEL-compatible rebuild · Open source, community-maintained CentOS | The CentOS Project | - | Reported | top500.org |
The read
Petrobras's seismic-imaging supercomputer at Vargem Grande in Rio de Janeiro, built by Atos with Supermicro A+ 4124GO-NART+ GPU servers: 2,016 NVIDIA A100 80GB GPUs, AMD EPYC 7513 CPUs, HDR InfiniBand and 678 TB of memory. It was billed as the largest supercomputer in Latin America at launch and debuted 33rd on the November 2022 TOP500 (117th by June 2026). Atos and NVIDIA quote almost 21 PFlop/s peak while TOP500 gives 42.00 PFlop/s Rpeak, so no peak is recorded and the conflict is logged. The exact site coordinates were not found in a page we fetched, so the system is left unlocated. An NVIDIA Brazil post dates it as operational on 9 January 2023, but the go-live date is not stated firmly, so it is blank. The A100 edge is verified by TOP500 and NVIDIA; Rmax, core and power figures come from TOP500 alone, so the system tier is reported.
Timeline
- 2022-11 List debut at #33 (November 2022) source
Public list appearances
Pointers only: rank and edition, linking to the canonical entry. We do not reproduce list tables. See sourcing policy.
Change history
- 2026-10-01 Added Pegasus (Petrobras)
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 |
|---|---|---|
| top500.org | does not name this system | AMD EPYC 7513 (Milan), NVIDIA InfiniBand HDR, CentOS |
| nextplatform.com | names system only | - |
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.