Presenter: Sneha Latha Kottapalli, Senior Systems Software Engineer, NVIDIA
Date: November 7, 2018
Optimizing software applications is an iterative process that involves profiling to identify the performance bottlenecks and tuning the application to address the bottlenecks until the best performance is achieved. When optimizing for heterogeneous systems with multiple CPUs and GPUs such as NVIDIA DGX or workstations, independent CPU profilers and GPU profilers are insufficient to understand the interactions between various system resources. NVIDIA Nsight Systems is a system-wide performance analysis tool that provides a holistic view of how the application is utilizing the system’s compute units. It helps developers to quickly identify the major performance bottlenecks and optimize their application performance. In this webinar, I will provide an overview of Nsight Systems and highlight its unique features. We will look at a few examples that demonstrate the most common performance issues.
Target audience: The webinar is intended for researchers, developers, and data scientists who are interested in optimizing their software for heterogeneous systems with multiple CPUs and GPUs.
Prerequisites: This webinar assumes basic knowledge of CPU, GPU, and operating system (OS).
Software Availability: Free to download from http://developer.nvidia.com/nsight-systems after registering with the NVIDIA Developer program. Note that the registration is free too.
Software Requirements: Please see “System Requirements” and “Release notes” pages in the product documentation https://docs.nvidia.com/nsight-systems/index.html
Training and reference materials:
- Nsight Systems documentation is available online at https://docs.nvidia.com/nsight-systems/index.html
- Blog post on Nsight Systems https://devblogs.nvidia.com/nsight-systems-exposes-gpu-optimization
- GTC 2018 talk introducing Nsight Systems http://on-demand-gtc.gputechconf.com/gtc-quicklink/MmmcS4
- User forum https://devtalk.nvidia.com/default/board/308/nsight-systems
Sneha Latha Kottapalli is a Senior Systems Software Engineer at NVIDIA, Champaign. Her team develops NVIDIA Nsight Systems, a system-wide performance analysis tool that helps optimize software applications for multi-GPU compute hardware. Sneha obtained her M.S. degree in Electrical Engineering from Clemson University in 2013 and a B.Tech degree in Electronics and Communication Engineering from National Institute of Technology Karnataka (NITK), Surathkal in 2009. Prior to joining NVIDIA, she was a software engineer at Microsoft (2013-2014) and at Intel (2014-2016).