Open data
Verified liquid-cooling deployment tracker
Liquid-cooling projects with a named operator, a location, a stage, and at least one source — separating what is operating from what was announced.
- Rows
- 9
- Columns
- 13
- Version
- 2026-08-09
- Generated
- 2026-09-24T05:08:37.204Z
- Identifier
- dcc:liquid-cooling-deployments
- Licence
- CC BY 4.0
9 rows. Verified liquid-cooling deployment tracker carries its methodology and limits inside the JSON file, so the file can be read without this page.
What you can answer with it
- Count operating capacity without counting announcements twice.
- See which architectures have moved past pilot in a given market.
- Find the projects a supplier can actually point to, and the ones it cannot.
How the rows were built
- Rows are the deployment records on this site. Each carries a stage taken from the most recent source, and the stage history is on the record.
- Stage is deliberately four values — operating, deploying, planned, pilot — because a finer scale would imply knowledge the sources do not contain.
- Capacity is quoted as the source states it. Where a project reports IT load and facility load differently, the record says which.
- A project with no source is not published, so it is not in this file.
The full editorial standard, including how sources are ranked and how project stages are labelled, is on the methodology page.
What it does not cover
- Operators disclose selectively. A market with few rows may be under-reported rather than under-built.
- Announced capacity frequently changes and is rarely restated publicly. Treat a figure as of its review date, not as current.
- Supplier attribution is limited to what a source names, so a project may involve suppliers this file does not list.
- Colocation projects are recorded at the facility, not per tenant, because tenant loads are almost never public.
PreviewFirst 9 of 9 rows
| Project | Operator | Location | Stage | Announced | Capacity | Rack density | Compute | Architecture | Named suppliers | Sources | Reviewed | Record |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DarkNX Ontario AI campus | DarkNX | Ontario, Canada | Planned | Nov. 17, 2025 | 300 MW planned; first phase two × 65 MW | Not disclosed | AI and high-density GPU workloads | Two-phase direct-to-chip cooling | Accelsius; Johnson Controls | 1 | Aug. 9, 2026 | Open record |
| Madhya Pradesh–Submer AI capacity program | Government of Madhya Pradesh and Submer | Madhya Pradesh, India | Planned | July 28, 2025 | Up to 1 GW under a memorandum of understanding | Not disclosed | AI-ready data centers | Direct-to-chip liquid cooling; Single-phase immersion cooling; Hybrid air and liquid cooling | Submer | 1 | Aug. 9, 2026 | Open record |
| Equinix Ashburn NeuCool deployment | Equinix | Ashburn, Virginia, United States | Operating | July 15, 2025 | IR80 demonstration system | Not disclosed | Co-innovation and customer validation | Two-phase direct-to-chip cooling | Accelsius | 1 | Aug. 9, 2026 | Open record |
| SoftBank–Foxconn H200 rack cooling pilot | SoftBank | Japan | Pilot | May 21, 2025 | One rack-scale demonstration | Not disclosed | NVIDIA H200 GPUs | Two-phase direct-to-chip cooling | ZutaCore | 1 | Aug. 9, 2026 | Open record |
| Stellium OCP immersion proof of concept | Stellium Data Centres | Newcastle, United Kingdom | Pilot | April 1, 2025 | Proof of concept | Not disclosed | Mitac Open Rack v3 servers with AMD CPUs | Single-phase immersion cooling | Submer | 1 | Aug. 9, 2026 | Open record |
| Submer Barcelona AI data center | Submer | Barcelona, Spain | Planned | Feb. 28, 2025 | 56 MW | 150 kW+ per rack or tank (design target) | AI infrastructure | Direct-to-chip liquid cooling; Single-phase immersion cooling; Hybrid air and liquid cooling | Submer | 1 | Aug. 9, 2026 | Open record |
| LiquidStack 300 MW U.S. multi-site order | Undisclosed U.S. data-center operator | Multiple sites, United States | Deploying | Jan. 6, 2025 | 300 MW aggregate CDU capacity | Not disclosed | AI-ready, high-density infrastructure | Direct-to-chip liquid cooling; Hybrid air and liquid cooling | LiquidStack | 1 | Aug. 9, 2026 | Open record |
| Shell Skybox Houston immersion cluster | Shell | Katy, Texas, United States | Operating | Jan. 1, 2024 | 1.2 MW | Up to 100 kW per 42U rack with warm water, company reported | Penguin Solutions high-performance computing cluster with AMD processors | Single-phase immersion cooling | GRC | 1 | Aug. 9, 2026 | Open record |
| TACC Lonestar6 immersion cooling | Texas Advanced Computing Center | Austin, Texas, United States | Operating | Jan. 31, 2022 | 280 kW immersed system | Not disclosed | Dell C6525 servers with AMD EPYC CPUs | Single-phase immersion cooling | GRC; Dell Technologies | 1 | Aug. 9, 2026 | Open record |
Data dictionary13 columns
projectProject- Project or site name as the sources give it.
operatorOperator- The party running the facility, not the tenant.
locationLocation- Stated to the granularity the source supports.
stageStage- Where the project is: announced, planned, pilot, deploying, or operating. Stage is not evidence strength.
announcedAnnounced- Date of the earliest public announcement on file.
capacityCapacity- As stated by the source; IT load and facility load are distinguished on the record.
densityRack density- Per-rack power as published, where published.
computeCompute- Accelerator or server platform named by the source.
architectureArchitecture- Cooling architectures in use, from the record's links.
suppliersNamed suppliers- Only suppliers a source names. Absence is not evidence of non-involvement.
sourcesSources- Count of cited sources on the underlying record.
reviewedReviewed- When a human last read the record this row came from.
recordRecord- Canonical URL of the record, so every row is traceable.
Versions and change log
This file is 2026-08-09, generated 2026-09-24T05:08:37.204Z from 9 records. No publish has changed this file since versions started being recorded, so the version quotes the most recent review across those records instead. The rows themselves are always regenerated from the records, so a correction reaches the file as soon as it reaches the record.
No published change has touched a record in this file yet. When one does, it appears here rather than only in the file.
Formats and licence
- CSVAvailable at /datasets/liquid-cooling-deployments/data.csv, RFC 4180 quoted.
- JSONAvailable at /datasets/liquid-cooling-deployments/data.json, with the same columns and row order.
- ParquetNot published. The row counts here do not justify it yet, and offering a format we do not regenerate would be worse than not offering it.
- APINo query API. The JSON file is the whole dataset, so fetching it is the API for now.
Licensed CC BY 4.0: Free to use, republish, and build on, including commercially, with attribution to DataCenterChill and a link to the dataset page. Read the licence.
Citing and correcting
Cite it as: DataCenterChill, Verified liquid-cooling deployment tracker (dcc:liquid-cooling-deployments), 2026-08-09, https://www.datacenterchill.com/datasets/liquid-cooling-deployments. Quoting a figure without the conditions in its row is the error this file exists to prevent, so keep them together.
A row that is wrong is a record that is wrong. Send the source and the correction reaches the record, this file, and every page connected to it.