Germany · under construction · ai training

HammerHAI

Operated by HLRS at HLRS, University of Stuttgart .

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	- HPE liquid-cooled NVIDIA GB200 NVL4 system	Hewlett Packard Enterprise HPENYSE	-	Reported	hlrs.de itdaily.com
Accelerator	- 215 NVIDIA GB200 NVL4 nodes with 860 Blackwell GPUs in total	NVIDIA NVDANASDAQ	860 reported	Reported	hlrs.de
CPU	NVIDIA Grace CPUs Armv9 (Neoverse V2) · 72C · 3.1 GHz · TSMC 4N Two NVIDIA Grace 72-core CPUs per GB200 NVL4 node	NVIDIA NVDANASDAQ	-	Reported	hlrs.de
Interconnect	NVIDIA InfiniBand XDR (Quantum-X800) Interconnect Fat tree or dragonfly+ · 800 Gb/s per port NVIDIA Quantum-X800 InfiniBand (800 Gbit/s) and Spectrum-X Ethernet	NVIDIA NVDANASDAQ	-	Reported	hlrs.de itdaily.com
Storage	VAST Data Platform Storage Disaggregated shared-everything 10 PB VAST Data DASE storage	VAST Data	-	Reported	hlrs.de

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	860 accelerators	2026-10	Design target	HLRS: 215 NVIDIA GB200 NVL4 nodes, 860 Blackwell GPUs. Under construction. Excludes 13 L4 and Axelera AIPU nodes.	hlrs.de

The read

The first new standalone supercomputer under the EuroHPC AI Factories initiative, contracted with HPE in March 2026 for the HammerHAI AI Factory at HLRS in Stuttgart. HLRS lists 215 NVIDIA GB200 NVL4 nodes (860 Blackwell GPUs, each node with two 72-core Grace CPUs), 9 Axelera AI nodes due in 2027, 13 nodes with NVIDIA L4 GPUs and 5 RTX A1000 nodes, 10 PB of VAST Data storage and Quantum-X800 InfiniBand, for up to 15 EFlop/s of peak AI inference. Delivery was scheduled for the second quarter of 2026 and operation for the second half of 2026; no confirmation of commissioning was found as of the status date, so it stays under construction. ITdaily independently confirms HPE, the GB200 NVL4 design, Quantum-X800 and the 15 EFlop/s figure but not the node count. Only AI-inference figures are published, so no HPL FLOPS are recorded.

Timeline

2026-03 Announced source

Site & facility

HLRS, University of Stuttgart

Location: Stuttgart, Germany
Cooling: Direct liquid cooling
Commissioned: 1996

hlrs.de

Change history

2026-10-01 Added HammerHAI (HLRS)

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
itdaily.com	names system and part	NVIDIA Grace, NVIDIA InfiniBand XDR (Quantum-X800)
hlrs.de	names system and part	NVIDIA Grace, NVIDIA InfiniBand XDR (Quantum-X800), VAST Data Platform

Checked 2026-10-09 by pnpm verify. Re-run it and the table changes with the web.