After clicking “Watch Now” you will be prompted to login or join.
Click “Watch Now” to login or join the NVIDIA Developer Program.
Memory Management on Modern GPU Architectures
Nikolay-Sakharnykh, NVIDIA | Max Katz, NVIDIA | Michael Delomre, NVIDIA | Lars Nyland, NVIDIA | Chirayu Garg, NVIDIA | Mark Hairgrove, NVIDIA
Come to our session to discuss any topics about memory management on GPU systems: new memory management APIs, profiling and optimizations for GPU memory subsystem, tips and tricks for managing data across multiple GPUs and CPUs, and compression techniques. Whether you're interested in the low-level details of the GPU architecture, or software heuristics in the driver, or guidelines and best practices for applications — we have the right experts from multiple NVIDIA teams you can connect with and chat about your topic.