Nimbus Cloud secures $120M to expand a GPU-as-a-service platform with edge deployments

Tech · 6 min read

Nimbus Cloud, a provider of on-demand GPU infrastructure, announced a $120 million growth round led by Tiger Global and plans to expand its footprint into edge locations for low-latency AI inference. Nimbus offers GPU instances optimized for model training and inference, plus orchestration tooling to manage multi-region deployments.

With the new funding, Nimbus will build regional edge PoPs, invest in scheduling and arbitration software for shared accelerators, and partner with telcos to place inference nodes closer to devices. The company also introduced a managed inference service aimed at companies deploying real-time features like AR filters and voice assistants.

Customers using Nimbus reported improved inference latency for global applications and easier scaling across regions. Nimbus highlighted a roadmap that includes hardware diversification, support for next-gen accelerators, and enhanced billing models tailored to bursty workloads.