红帽和AWS正在合作,在AWS的Trainium和Inferentia芯片上运行红帽的AI推断服务器,承诺比GPU价格性能提高40%,开发人员预览版将于2026年1月发布. Red Hat and AWS are teaming up to run Red Hat’s AI Inference Server on AWS’s Trainium and Inferentia chips, promising 40% better price-performance than GPUs, with a developer preview set for January 2026.

红帽子和AWS扩大了伙伴关系,在AWS的训练场和发酵芯片上运行红帽子的AI推论服务器,比传统的以GPU为基础的案例价格表现好40%。 Red Hat and AWS have expanded their partnership to run Red Hat’s AI Inference Server on AWS’s Trainium and Inferentia chips, offering up to 40% better price-performance than traditional GPU-based instances. 集成支持 OpenShift , 并包括一个新的 AWS 中子操作器、 Aseable 收藏和用于简化 AI 部署的 vLLM 插件 。 The integration supports OpenShift and includes a new AWS Neuron operator, Ansible collection, and vLLM plugin for streamlined AI deployment. 该解决方案定于2026年1月开发商预览,旨在提高企业AI的效率和可扩展性。 The solution, set for developer preview in January 2026, aims to boost enterprise AI efficiency and scalability.