COMPUTE ATLAS Supercomputer supply-chain graph
1153 systems 482 sites Sourced data

Supercomputers/Systems/Baidu Kunlun P800 Cluster

China · operational · mixed

Baidu Kunlun P800 Cluster

Operated by Baidu .

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
Accelerator -
30,000 Baidu Kunlun P800 chips
Baidu
30,000
reported
Reported finance.yahoo.comcaixinglobal.com

AI datacenter metrics

Figures beyond an accelerator count, each with who stated it. What can and cannot be compared across AI datacenters.

MetricValueAs ofBasisNoteSource
Accelerators 30,000 accelerators 2025-04 Operator stated Baidu said at its April 2025 developer conference it had lit a cluster of 30,000 P800 Kunlun chips (Reuters report). No power figure published. investing.com

The read

Baidu's cluster of its own Kunlun P800 AI chips. Reuters (via Yahoo Finance) reports CEO Robin Li announced on 24 April 2025 that a cluster of 30,000 P800 chips had been switched on (illuminated), able to train DeepSeek-like models with hundreds of billions of parameters. Caixin reports Baidu lit its first 10,000-card P800 cluster in February 2025, expanded it to 30,000 cards by April 2025 with most inference workloads running on it, and by 12 November 2025 had grown a separate 5,000-card training cluster to over 10,000 cards; the P800 clusters serve over 100 customers. Both figures are Baidu claims; no HPL result, node count, power or site is published, so the site is left unlocated. This entry tracks the 30,000-card cluster.

Timeline

Change history

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 sourceResultFound on the page
finance.yahoo.comcould not be fetched-
caixinglobal.comnames system only-

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

Further reading