DEPLOYDatabase

Company

Clockwork.io

Palo Alto company building fault-tolerance and observability software for GPU clusters (FleetLens, LinkPass, TorchPass) that keeps AI training, inference and…

Status
Private

Overview

Palo Alto company building fault-tolerance and observability software for GPU clusters (FleetLens, LinkPass, TorchPass) that keeps AI training, inference and reinforcement learning jobs running through hardware and network failures.

Reality check

Goodput claim (company-stated, Oct 2026)

SemiAnalysis's Dylan Patel is quoted saying TorchPass cut training goodput loss from 14% to under 3% for a neocloud in ClusterMAX testing.

Verified record

Verified deployments
None on file
Active incidents
None on file

Key facts

Funding (company-stated, Oct 2026)

Raised $31M on Oct 5 2026, co-led by Premji Invest, Wing Venture Capital and Seligman Ventures with NEA and e& Capital; $73M total.

Product

A GPU fault-tolerance suite: LinkPass reroutes traffic around failed links; TorchPass moves work off failing GPUs.

TorchPass updates (company-stated, Oct 2026)

TorchPass is adding multi-node platform snapshots and fast asynchronous checkpoints.

Customers (company-stated, Oct 2026)

In production at LinkedIn; Together AI offers TorchPass as a service on its GPU Clusters; WhiteFiber is expanding its use.

LinkedIn result (company-stated)

LinkedIn says the suite prevents tens of thousands of GPU-hours of downtime a month.

Data & sources

Press releases

1

Web sources

2

3 sources backing this record.View all →

Safety record

No incidents on record for Clockwork.io.

Only active incidents are counted. Retracted incidents are excluded from this summary but remain reachable at their canonical URLs.

Full safety record: incidents, sourcing, and exposure data →

Recent coverage

Clockwork.io in third-party press

Mentioned on (1)

Registry pages whose text names Clockwork.io.

Companies (1)

WhiteFiber