DigitalOcean, a scalable cloud infrastructure provider for enterprises, announced an expansion of its inference capabilities in partnership with Nvidia as it shifts toward building an AI Factory to support development of autonomous agents.
“Traditional hyperscalers offer fragmented, overly complex environments. Our AI Factory removes that friction, giving builders a direct, simplified path to scale,” Vinay Kumar, DigitalOcean’s chief product and technology officer, wrote in a company blog.
DigitalOcean is launching a new data center in Richmond to support its AI and inference services. It will feature Nvidia HGXTM B300 systems and support massive-scale model deployments.
DigitalOcean is also integrating advanced open models into its Nvidia-native stack, deployed on its own high-performance infrastructure. It introduced Nvidia Dynamo 1.0 for scaling production inference, an agent building toolkit and new models for agentic workflows. The company is collaborating with Nvidia on an open-source stack known as NVIDIA NemoClaw to enable securely running always-on assistants through open-source agent OpenClaw with a single command.
