This course provides hands-on and conceptual grounding in high-performance GPU computing on Marlowe. Aimed at researchers from diverse disciplines, it introduces performance optimization strategies, resource scaling, and reproducibility best practices. Stanford and NVIDIA engineers will lead specialized sessions on multi-GPU scaling and profiling using their ecosystem of tools.
1 units · Satisfactory/No Credit
This course provides hands-on and conceptual grounding in high-performance GPU computing on Marlowe. Aimed at researchers from diverse disciplines, it introduces performance optimization strategies, resource scaling, and reproducibility best practices. Stanford and NVIDIA engineers will lead specialized sessions on multi-GPU scaling and profiling using their ecosystem of tools.
Offered in Autumn 2025 at Stanford University.