The open-source inferencing software increases throughput and reduces the cost of LLM token generation, the chipmaker said.

Ian Buck, the head of hyperscale and high-performance computing at NVIDIA, explained the prowess of NVIDIA Dynamo. He said, ...