SambaNova

SambaNova builds full-stack AI platforms with specialized chips for fast inference, training, and agentic AI in data centers and clouds.

SambaNova develops AI hardware like RDU chips and integrated software for running large-scale models with high speed and energy efficiency. These systems handle inference on massive models up to trillions of parameters, support quick model switching for agentic workflows, and work in data centers or cloud setups. The platform runs open-source models such as DeepSeek, Llama, and gpt-oss with OpenAI-compatible APIs, helping enterprises deploy AI without heavy power use.