NVIDIA: GPU Acceleration Workshop

New

This course is being delivered in collaboration with the UIS and NVIDIA

According to NVIDIA, accelerated computing is the future for HPC and AI. Effective accelerated computing applications can only be built on a complete solution stack comprising hardware, programming models, libraries, system software, developer tools and domain expertise. NVIDIA’s CUDA is one of the leading platforms and this workshop aims to introduce researchers using large scale computing resources to the opportunities available in GPU-accelerated computing.

• Please note that if you are not eligible for a University of Cambridge Raven account you will need to book by completing this form.

• All University (departments and colleges) Computational Researchers interested in learning more about GPU computing, specifically for increased performance on HPC systems
• EPSRC Tier2 and DiRAC users may also attend
• Further details regarding eligibility criteria are available

• Familiarity with the software stacks for high-performance computing and artificial intelligence.

Number of sessions: 1

• Accelerated computing is a full stack problem
• The NVIDIA HPC platform
• The CUDA platform
• Software architecture trends
• CUDA platform deployment
• CUDA tools and libraries
• HPC + X
• Discussion and wrap-up

• For further information on the course please contact js947@cam.ac.uk