Topic: scientific computation

Research computing platforms and methods that improve the efficiency of machine-learning workloads. The appointment contributes to RIKEN’s AI for Science Platform and the optimisation of scientific computation.

Develop high-performance artificial-intelligence systems that accelerate scientific computation. Research includes AI-assisted algorithms, scalable model training and computational methods.

Shayan Majidy examines how AI can help build and use quantum computers. Algorithm-specific error-correcting codes offer substantial improvements for targeted tasks, with machine learning, SAT methods and tensor networks extending code design. The talk also presents an experimentally realised universal fault-tolerant gate set and the use of learning to optimise its bottlenecks, then considers future integration of large quantum processors and AI for scientific computation.