Supercomputers/Systems/Govorun
Russia · upgraded · hpc
Govorun
Also known as HybriLIT Govorun supercomputer
Operated by JINR at JINR Information Technology Laboratory .
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
- -
- Rpeak
- 2.20 PFlop/s
- Rmax ÷ Rpeak
- -
- Cores
- -
- 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 | - RSC Tornado hot-water liquid-cooled compute nodes and RSC Exastream AI servers | RSC Group | - | Reported | rscgroup.ruhlit.jinr.ru |
| Accelerator | NVIDIA H100 80GB PCIe Accelerators CUDA (Hopper) NVIDIA H100 PCIe 80GB GPUs in two RSC Exastream AI servers of eight each | NVIDIA | 16 reported | Reported | hlit.jinr.rujinr.ru |
| Accelerator | NVIDIA Tesla V100 Accelerators CUDA (Volta GV100) · 5,120C · 1.53 GHz · 12 nm · 300 W NVIDIA V100 GPUs in five DGX-1 servers of eight each | NVIDIA | 40 reported | Reported | hlit.jinr.ru |
| Accelerator | - NVIDIA A100 GPUs in five Niagara R4206SG servers of eight each (A100 variant not stated) | NVIDIA | 40 reported | Reported | hlit.jinr.ru |
| CPU | Intel Xeon Platinum 8280 CPUs x86-64 · 28C · 2.7 GHz · 14 nm · 205 W Intel Xeon Platinum 8280 28C nodes among CPU nodes that also include Xeon Platinum 8268, Xeon Platinum 8368Q and Xeon Phi 7190 (HybriLIT page) | Intel | - | Reported | hlit.jinr.ru |
| Interconnect | Omni-Path Architecture Interconnect Fat tree · 100 Gb/s per port · About 1 microsecond Intel Omni-Path 100 Gbps fat-tree fabric with liquid-cooled 48-port switches | Intel | - | Reported | rscgroup.ruhlit.jinr.ru |
| Cooling | Direct liquid cooling, warm water Cooling Direct-to-chip cold plates, facility water loop · Removes 90+ percent of node heat to liquid 100 percent hot-water direct liquid cooling with RSC CDU-HM 400 distribution unit | RSC Group | - | Reported | rscgroup.ruhlit.jinr.ru |
The read
The Joint Institute for Nuclear Research's heterogeneous supercomputer at its Laboratory of Information Technologies in Dubna, built by RSC with Tornado hot-water liquid-cooled nodes. RSC says it was deployed between November 2017 and March 2018 on Xeon Platinum 8268, Xeon Gold 6154 and Xeon Phi 7290 nodes with Intel Omni-Path. The HybriLIT page now describes a CPU part of 800 TFlops double precision and a GPU part of 1.4 PFlops double precision (58 PFlops half precision) made of 16 NVIDIA H100, 40 A100 and 40 V100 GPUs, for a stated total peak of 2.2 PFlops double precision and 10.6 PB of storage. A JINR news item of 8 July 2025 reports the upgrade that added two nodes of eight H100 PCIe 80GB GPUs each. The 2.2 PFlops is an operator-stated peak, not a Linpack run, and the system is not on the TOP500 list; the two cited pages are from the operator and its integrator, so the tier is reported.
Timeline
- 2018 First operational source
Site & facility
JINR Information Technology Laboratory
- Location
- Dubna, Russia
- Cooling
- Direct liquid cooling (hot water)
Change history
- 2026-10-02 Added Govorun, JINR's hot-water-cooled supercomputer
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 |
|---|---|---|
| hlit.jinr.ru | names system, part and count | Intel Xeon Platinum 8280, NVIDIA H100 80GB PCIe, Omni-Path Architecture, Direct liquid cooling, warm water · counts: 16, 40 |
| rscgroup.ru | names system, part and count | Intel Xeon Platinum 8280, Omni-Path Architecture, Direct liquid cooling, warm water · counts: 16, 40 |
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.