Beginning of dialog window. Escape will cancel and close the window.
End of dialog window.
详情
字幕
Demystify CUDA Debugging and Performance with Powerful Developer Tools
, Technical Product Manager, NVIDIA
CUDA applications running on NVIDIA GPUs are some of the top performing workloads running today. Unlocking this performance doesn't require expert knowledge of the hardware as long as you have the right tools for the job. We'll start with a brief overview of the tools available for free, and then dive deep on the latest features that provide even more insights into application behavior and performance. These include new optimization recipes for cluster-scale GPU workloads, a timeline for key kernel performance metrics, and source comparison features to show exactly how and why source modifications impact performance. Additionally, debugging tools have added several improvements to make bug detection and correction easier. Learn how to unlock the performance potential of CUDA applications.