Supercomputers/Systems/Horizon
USA · under construction · mixed
Horizon
Operated by Texas Advanced Computing Center at TACC Horizon data center (Round Rock) .
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
- -
- 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 | - Dell servers | Dell Technologies | - | Reported | tacc.utexas.edu |
| Accelerator | NVIDIA GB200 NVL72 Accelerators Arm CPU + CUDA GPU · TSMC 4NP NVIDIA Grace Blackwell servers | NVIDIA | - | Reported | tacc.utexas.edu |
| Interconnect | NVIDIA InfiniBand NDR (Quantum-2) Interconnect Fat tree or dragonfly+ · 400 Gb/s per port · Under 0.6 microseconds NVIDIA Quantum-2 InfiniBand | NVIDIA | - | Reported | tacc.utexas.edu |
| Storage | VAST Data Platform Storage Disaggregated shared-everything 400 PB all-solid-state storage, VAST Data | VAST Data | - | Reported | tacc.utexas.edu |
AI datacenter metrics
Figures beyond an accelerator count, each with who stated it. What can and cannot be compared across AI datacenters.
| Metric | Value | As of | Basis | Note | Source |
|---|---|---|---|---|---|
| Accelerators | 4,000 accelerators | 2026-10 | Design target | TACC: Grace Blackwell and Vera servers with 4,000 GPUs. Planned system. Both pages are TACC. | tacc.utexas.edutacc.utexas.edu |
| Stated investment | 0.46 USD bn | 2024-07 | Operator stated | NSF award 2323116, Construction for the Leadership Class Computing Facility: $457.4M estimated total. Covers Horizon plus facility, visitor centre and partners. | api.nsf.gov |
| Facility power | 20 MW | 2026-10 | Design target | TACC: new 15-20 MW liquid-cooled data centre in Round Rock. 20 is the top of the stated range; planned capacity, not a draw. | tacc.utexas.edu |
The read
TACC's NSF Leadership-Class Computing Facility system, being installed in a Sabey data center in Round Rock with Dell, NVIDIA Grace Blackwell and Vera servers, Quantum-2 InfiniBand and 400 PB of solid-state storage from VAST Data. TACC's two pages disagree: the November 2025 post says 300 PFlop/s and production in spring 2026, while the systems page says 360 PFlop/s and phase one in fall 2026. Neither FLOPS figure nor either date is recorded. Both pages are TACC's, so the tier is reported.
Timeline
- 2024-07 Announced source
Site & facility
TACC Horizon data center (Round Rock)
- Location
- Round Rock, TX, USA
- Commissioned
- 2026
Where sources disagree
We record conflicts instead of picking a winner quietly. All open questions.
- Peak performance: 300 PFlop/s (November 2025 post) against 360 PFlop/s (systems page). Both pages are TACC; neither figure is recorded.
- Production date: Spring 2026 (November 2025 post) against Phase 1 in fall 2026 (systems page). The later page is treated as current, but no date is recorded.
Change history
- 2026-09-20 Added from operator sources
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 |
|---|---|---|
| tacc.utexas.edu | names system and part | NVIDIA InfiniBand NDR (Quantum-2), VAST Data Platform |
| tacc.utexas.edu | names system and part | NVIDIA InfiniBand NDR (Quantum-2), VAST Data Platform |
Checked 2026-10-09 by pnpm verify. Re-run it and the table changes with the web.
Related news
- Dell and NVIDIA detail Horizon, the NSF's Frontera successor at TACC · Dell TechnologiesHorizon pairs unreleased NVIDIA Grace Blackwell and Vera parts with Dell's direct liquid cooled IRSS racks and Quantum-X800 InfiniBand, targeted at roughly 10x Frontera's simulation throughput. NSF funding uncertainty flagged earlier in 2025 did not stop the build, and early user access is slated for spring 2026, which is worth checking against whatever actually ships.
Further reading
Inside a node
Why accelerators exist, what the memory hierarchy costs, and how the CPU and accelerator merged onto one package.
Operating a supercomputer
Acceptance testing, why "installed" and "in production" are six months apart, and the machine lifecycle.
The interconnect
Topologies, why latency and tail behaviour matter more than bandwidth, and how the fabric market consolidated.