akjus/openi-octopus

OPENI is an open-source cluster management tool and resource scheduling platform designed for AI jobs, particularly deep learning workloads, in GPU clusters. It provides a microservices-based architecture with support for major deep learning frameworks like TensorFlow and CNTK, while offering fine-grained GPU scheduling and topology-aware job placement. The platform is primarily intended for academic and research purposes, allowing users to evaluate various AI-related research ideas, from scheduling mechanisms to new hardware architectures.