reported tenant handoff to newly spun-up nodes after a one- or two-node failure
~90 secobserved
| Value kind | observed — Reported measurements, counts, and specifications keep the precision and scope stated by their source; an exact specification is not treated as a range. |
|---|---|
| As of | 2026 |
| Source | SemiAnalysis AI Neocloud Playbook / ClusterMAX · Failure handling discussion: top GPU clouds spin up replacement nodes within 90 seconds |
| Review | checking…review by 2026-09-06 · standard cadence |
| Recorded changes | last 2026-06-29 |
| Claim id | automated-node-replacement-on-a-best-in-class |
Where the guide uses it
← Full numbers register — every date-stamped figure in the guide, with revision history.