Project requirement: A research institute in Shandong focuses on the forefront of artificial intelligence, aiming to break through basic mechanisms, models, and algorithm bottlenecks, and overcome new generation common key technologies. It requires a computing power cluster to support deep learning research, and a unified managem-ent and scheduling platform to integrate backend computing resources.

Solution: Multi node Turing evolutionary computing server l Management server l Turing distributed storage l Computing network: 100G IB networking l Cloud platform software: Artificial intelligence cloud platform.

Customer Benefits: Carry out research on key technologies such as machine learning, computer vision, natural language processing, knowledge expression and reasoning, and support application research in industries such as intelligent healthcare, smart homes, and intelligent transportation. Solve the bottleneck problem of computing power demand for models and algorithms through computing power servers. Users can directly allocate computing resources, store resources, and distribute them to multiple users on the software platform. The software platform integrates mainstream AI frameworks and acceleration software libraries such as Rapids for machine learning and data science, making it convenient for users to do AI related research work.