This technology is developed to meet enterprise AI deployment needs. Based on the Kubernetes container orchestration framework (high efficiency, low cost, fast deployment, high concurrency, supporting containerized independent service management), it is compatible with mainstream deep learning framework container images such as TensorFlow and PyTorch. Leveraging this, a shared GPU distributed container scheduling system is developed, building an intelligent machine learning platform for AI deployment.

Technology provider:Nanjing University of Posts and Telecommunications