Clockwork.io Raises $31 Million for AI Infrastructure Resilience

Clockwork.io raised $31 million and reported production deployments at LinkedIn and Together AI, the company said in a press release. Premji Invest, Wing Venture Capital and Seligman Ventures jointly led the round, with participation from NEA and e& Capital.

The financing brings Clockwork.io's total funding to $73 million. The company plans to use the funds to deploy its fault tolerance software across AI training, inference and reinforcement learning workloads, and to expand distribution through cloud partners.

LinkedIn uses Clockwork.io's LinkPass software across its AI infrastructure fleet. The company says the system redirects traffic around network failures and prevents tens of thousands of hours of GPU downtime each month. Together AI offers TorchPass as a service on its GPU clusters, while WhiteFiber is expanding Clockwork.io software across its global GPU service network.

Clockwork.io also introduced two TorchPass features. Multi node platform snapshots save the state of distributed AI jobs without code changes, while asynchronous application checkpoints transfer updated model weights to inference replicas during reinforcement learning.