View timeline for applications that use CUDA Dynamic Parallelism including both host-launched and device-launched kernels and the parent-child relationship between kernels. Use the command line profiler using environment variables to collect data from multiple systems and analyze the results in Visual Profiler Analyze data collected from remote systems.Compare results across multiple sessionsĬonfirm performance improvements by comparing against previous sessions.Gain low-level insights by looking at performance metrics collected directly from GPU hardware counters and software instrumentation. View all memory transfers, kernel launches, and other API functions on the same timeline View CUDA activity occurring on both CPU and GPU in a unified time line, including CUDA API calls, memory transfers and CUDA launches. Perform automated analysis of your application to identify performance bottlenecks and get optimization suggestions that can be used to improve performance Quickly identify potential performance bottleneck issues in your applications using highly configurable tables and graphical views You may download this and other macOS tools using the button below. While there are no tools which use macOS as a target environment, NVIDIA is making a macOS host version of Visual Profiler available from which you can launch profiling sessions on supported target platforms. Note that NVIDIA® CUDA Toolkit 11.0 (and later) no longer supports development or running applications on macOS.
The NVIDIA Visual Profiler is available as part of the CUDA Toolkit. First introduced in 2008, Visual Profiler supports all 350 million+ CUDA capable NVIDIA GPUs shipped since 2006 on Linux, Mac OS X, and Windows. The NVIDIA Visual Profiler is a cross-platform performance profiling tool that delivers developers vital feedback for optimizing CUDA C/C++ applications.
0 kommentar(er)
