CoreWeave has announced a new collaboration with NVIDIA to offer Training Cluster as a Service, providing access to powerful AI training infrastructure on demand.
“This service enables organizations to scale their AI training workloads without the upfront costs of building and maintaining their own clusters.”
The partnership combines CoreWeave's cloud platform with NVIDIA's latest GPU technology, including H100 and upcoming B200 GPUs, interconnected with high-speed networking. Customers can rent pre-configured clusters optimized for training large language models, computer vision models, and other deep learning applications.
This move is part of a broader trend where cloud providers are offering specialized hardware-as-a-service to meet the surging demand for AI compute power.