Roofline and NVIDIA Ampere GPU Architecture Analysis

This demo shows the latest CUDA kernel analysis capabilities in NVIDIA Nsight Compute, including the popular Roofline Analysis Method and a new feature for the NVIDIA Ampere GPU Architecture. Specifically, we'll demonstrate profiling the hardware-supported asynchronous data copy feature, which can boost the performance of workloads that are able to take advantage of it. #NsightCompute #RooflineAnalysis #SC20
Learn more about Nsight Compute, the entire family of Nsight Tools, and Download today. https://nvda.ws/3dPcplW
https://nvda.ws/35nFbX3
https://nvda.ws/3jvjdXj"