------------------------------------------------------------------------------ NVIDIA Compute Visual Profiler Windows XP, Vista, Server 2008, and Windows 7 Release Notes Version 3.1 ------------------------------------------------------------------------------ PLEASE REFER EULA.txt FOR THE LICENSE AGREEMENT FOR USING NVIDIA SOFTWARE. Please refer Changelog.txt for changes with respect to the previous version. FILES IN THE RELEASE: -------------------- * computeprof\bin\computeprof.exe : Compute Visual Profiler executable for Windows * computeprof\bin\*.dll : Qt DLLs * computeprof\bin\vcredist_x86.exe : Microsoft Visual C++ 2008 Redistributable Package (x86) distributed with Microsoft Visual Studio 2008. * computeprof\projects : Directory containing sample profiler projects * computeprof\doc : Directory containing files for online help and user documentation. SYSTEM REQUIREMENTS ------------------- * CUDA-enabled GPU See http://www.nvidia.com/object/cuda_learn_products.html * NVIDIA Driver * NVIDIA CUDA Toolkit INSTALLATION AND SETUP --------------------- The installation is part of the CUDA toolkit installation. The files are installed under "\computeprof" where is the directory under which the CUDA Toolkit is installed. If you do not have Microsoft Visual C++ 2008 or Microsoft Visual C++ 2008 Redistributable Package installed you will need to install the Microsoft Visual C++ 2008 Redistributable Package by running vcredist_x86.exe available under the "\computeprof\bin" directory. If the correct versions of Microsoft Visual C++ DLLs are not available when running Visual Profiler you will get the error: "Application failed to start because side-by-side configuration is incorrect." To resolve this, follow the steps mentioned above. RUNNING THE Compute Visual Profiler -------------------------------- From the Windows Start menu: Start->All Programs->NVIDIA Corporation->CUDA Toolkit->Compute Visual Profiler->Compute Visual Profiler Refer the Compute Visual Profiler computeprof.html document for more information. KNOWN ISSUES ------------ 1) Following are some issues related to profiler counters: . "warp serialize" counter for GPUs with compute capability 1.x is known to give incorrect and high values for some cases. . "divergent branch" counter for GPUs with compute capability 2.0 is known to give an incorrect value zero for some cases. . For GPUs with compute capability 2.0 the "instructions issued" and "instructions executed" counter values are incorrect for some cases. 2) The Compute Visual Profiler does not detect the CUDA version and does not disable any incompatible features. 3) If some OpenCL resources (contexts, events, etc.) are not released in the program, the profiler output may be incomplete or empty and Visual profiler will report the message ‘Error in reading profiler output'. The program needs to be modified to properly free up all OpenCL resources before termination. 4) Currently there is no way to pass keyboard input to a console-based CUDA/OpenCL program which is run from the Compute Visual Profiler. Therefore you must use the command line argument "--noprompt" to run most of the CUDA/OpenCL SDK samples inside the Compute Visual Profiler.