bwForCluster Helix

Also known as Helix; bwForCluster Helix Heidelberg

Operated by Heidelberg University at Heidelberg University Computing Centre (URZ) .

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	- MEGWARE computing cluster (storage supplied by ABC Systems)	MEGWARE	-	Reported	urz.uni-heidelberg.de
Accelerator	- NVIDIA A100, A40 and H200 GPUs: 55 four-GPU nodes and 7 eight-GPU nodes (about 200 Ampere GPUs at launch)	NVIDIA NVDANASDAQ	-	Reported	wiki.bwhpc.de urz.uni-heidelberg.de
CPU	AMD EPYC 7513 (Milan) CPUs x86-64 · 32C · 2.6 GHz · 7 nm · 200 W 355 standard plus 15 fat nodes with two AMD EPYC 7513 (64 cores per node) (quantity is nodes)	AMD AMDNASDAQ	370 reported	Reported	wiki.bwhpc.de urz.uni-heidelberg.de
Interconnect	NVIDIA InfiniBand HDR Interconnect Fat tree or dragonfly+ · 200 Gb/s per port · About 0.6 microseconds Non-blocking NVIDIA Mellanox InfiniBand HDR, HDR100 on standard nodes up to 4x HDR200 on GPU nodes	NVIDIA NVDANASDAQ	-	Reported	urz.uni-heidelberg.de wiki.bwhpc.de
Storage	- IBM Spectrum Scale (Storage Scale) file system, 11 PB with about 800 TB flash	IBM IBMNYSE	-	Reported	urz.uni-heidelberg.de wiki.bwhpc.de

The read

The state-level cluster for life sciences, natural sciences and computational humanities of Baden-Wuerttemberg, built by MEGWARE for Heidelberg University's computing centre (storage by ABC Systems) after a procurement that closed in June 2021, and live on 24 August 2022 as successor to bwForCluster MLS&WISO. Heidelberg describes about 20,000 AMD EPYC Milan cores, 100 TB of RAM, 200 NVIDIA Ampere GPUs (A100 and A40), non-blocking NVIDIA Mellanox InfiniBand HDR and 11 PB of IBM Spectrum Scale storage including 800 TB of flash. The bwHPC wiki now lists 355 standard and 15 fat nodes with two EPYC 7513 (64 cores per node), 55 four-GPU nodes with A40 or A100 and 7 eight-GPU nodes with A100 or H200, with interconnect from HDR100 to 4x HDR200. No HPL or peak figure is published in these sources, and the stated core count is approximate, so the cores field is blank.

Timeline

2021-06 Announced source
2022-08-24 First operational source

Site & facility

Heidelberg University Computing Centre (URZ)

Location: Heidelberg, Germany
Commissioned: 2022

urz.uni-heidelberg.de

Change history

2026-10-04 Added bwForCluster Helix, Heidelberg

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
wiki.bwhpc.de	names system and part	AMD EPYC 7513 (Milan), NVIDIA InfiniBand HDR
urz.uni-heidelberg.de	names system and part	AMD EPYC 7513 (Milan), NVIDIA InfiniBand HDR

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