华为联合高校发布并开源AI容器技术 助力算力利用效率提升

Core Insights - Huawei officially launched the AI container technology Flex:ai at the 2025 AI Container Application Implementation and Development Forum, aiming to address the low utilization of computing resources in the AI industry [1] Group 1: Technology Development - Flex:ai is a software for pooling and scheduling XPU resources, built on the Kubernetes container orchestration platform, which aims to enhance the utilization of AI workloads and computing resources [1] - The technology integrates research capabilities from three major universities and Huawei, achieving breakthroughs in three core technologies [1] Group 2: Resource Optimization - The XPU pooling framework developed in collaboration with Shanghai Jiao Tong University allows a single GPU or NPU card to be divided into multiple virtual computing units, improving overall computing utilization by 30% in scenarios where small AI models are trained [2] - A cross-node virtualization technology developed with Xiamen University aggregates idle XPU resources within a cluster to form a "shared computing pool," enabling general servers to forward AI workloads to remote GPU/NPU cards [2] Group 3: Intelligent Scheduling - The Hi Scheduler, developed with Xi'an Jiaotong University, provides intelligent scheduling for heterogeneous computing resources, ensuring optimal resource allocation for AI workloads even under fluctuating loads [3] - The comprehensive open-source nature of Flex:ai will allow developers from various sectors to access core technological capabilities, promoting the establishment of standardized solutions for efficient computing resource utilization in the global AI industry [3]