最新报道:Ionstream.ai, a GPU bare-metal cloud provider, is supporting the open-source SGLang framework by donating GPU credits on NVIDIA B200 GPUs to enhance tokenization efficiency. This collaboration aims to boost throughput, reduce latency, optimize memory usage, and improve cost-effectiveness for AI inference workloads, leveraging ionstream's 25 years of data center expertise and SGLang's innovations in language model serving.