Errata
Known Issues
General CUDA
When the default CUDA 5.0 Windows installer option to silently install the NVIDIA display driver is used, an error message like "display driver has failed to install" may be displayed for certain hardware configurations. If this error message occurs, the installation can be completed by installing the display driver separately using the setup.exe saved under C:\NVIDIA\DisplayDriver\....
In certain hardware configurations, the CUDA 5.0 installer on Windows may fail to install the display driver. This failure occurs when the user disables silent installation of the display driver and instead chooses to interactively select the components of the display driver from the installer UI that appears after the CUDA toolkit and samples are installed. If the UI for interactive selection of the display driver components fails to appear, please reinstall just the display driver by running setup.exe saved under C:\NVIDIA\DisplayDriver\....
On GPUs that are not in Tesla Compute Cluster (TCC) mode under Windows, CUDA streams may not achieve as much concurrency as they did in prior releases.
CUDA Libraries
The cublasgeam() routine provides undefined results if the pointer mode is set to CUBLAS_POINTER_MODE_DEVICE and the value pointed to by alpha is zero. There are two possible workarounds for this issue. The first is to use CUBLAS_PONTER_MODE_HOST instead of CUBLAS_POINTER_MODE_DEVICE, but this may require an extra device-to-host memory copy, depending on the situation. The second is to swap the (transa, alpha, A, lda) parameters with the (transb, beta, B, ldb) parameters, which would make the value pointed to by beta equal to 0.
The routine cublasCsyrk() may produce incorrect results on GPUs that implement the sm_30 architecture when the size of matrix parameter A exceeds (128M - 512) total elements.
The CUSPARSE library routines csrsv_analysis(), csrsv_solve(), csrsm_analysis(), and csrsm_solve() support the CUSPARSE_MATRIX_TYPE_GENERAL matrix type in addition to the supported matrix types already listed in the documentation.
CUDA Tools
The hardware counter (event) values may be incorrect in some cases on GPUs with compute capability (SM type) 3.5. Incorrect event values also result in incorrect metric values. These errors are more likely to occur when the same GPU is used for display and compute, or when other graphics applications are running simultaneously on the GPU.
Beginning with CUDA 5.0, the ptxas portion of the compiler generates a warning when the command line option "-abi=no" is used that indicates the option may be deprecated in a future release.
The current 5.0 linker will not support JIT to future architectures; objects will have to re-linked for each architecture.
Source-level analysis in NVIDIA Nsight Eclipse Edition and NVIDIA Visual Profiler is not available for kernels accessed through static function pointers.
Release Highlights
CUDA Dynamic Parallelism allows __global__ and __device__ functions running on the GPU to launch kernels using the familiar <<< >>> syntax and to directly call CUDA Runtime API routines (previously this ability was only available from __host__ functions).
All __device__ functions can now be separately compiled and linked using NVCC. This allows creation of closed-source static libraries of __device__ functions and the ability for these libraries to call user-defined __device__ callback functions. The linker support is considered to be a BETA feature in this release.
Nsight Eclipse Edition for Linux and Mac OS is an integrated development environment UI that allows developing, debugging, and optimizing CUDA code.
A new command-line profiler, nvprof, provides summary information about where applications spend the most time, so that optimization efforts can be properly ocused.
This release contains the following:
- NVIDIA CUDA Toolkit documentation
- NVIDIA CUDA compiler (nvcc) and supporting tools
- NVIDIA CUDA runtime libraries
- NVIDIA CUDA-GDB debugger
- NVIDIA CUDA-MEMCHECK
- NVIDIA Visual Profiler, nvprof, and command-line profiler
- NVIDIA Nsight Eclipse Edition
- NVIDIA CUBLAS, CUFFT, CUSPARSE, CURAND, Thrust, and
- NVIDIA Performance Primitives (NPP) libraries
Release notes in full:
http://docs.nvidia.com/cuda-toolkit-release-notes/index.html
Download & Instruction:
http://docs.nvidia.com/
Known Issues
General CUDA
When the default CUDA 5.0 Windows installer option to silently install the NVIDIA display driver is used, an error message like "display driver has failed to install" may be displayed for certain hardware configurations. If this error message occurs, the installation can be completed by installing the display driver separately using the setup.exe saved under C:\NVIDIA\DisplayDriver\....
In certain hardware configurations, the CUDA 5.0 installer on Windows may fail to install the display driver. This failure occurs when the user disables silent installation of the display driver and instead chooses to interactively select the components of the display driver from the installer UI that appears after the CUDA toolkit and samples are installed. If the UI for interactive selection of the display driver components fails to appear, please reinstall just the display driver by running setup.exe saved under C:\NVIDIA\DisplayDriver\....
On GPUs that are not in Tesla Compute Cluster (TCC) mode under Windows, CUDA streams may not achieve as much concurrency as they did in prior releases.
CUDA Libraries
The cublasgeam() routine provides undefined results if the pointer mode is set to CUBLAS_POINTER_MODE_DEVICE and the value pointed to by alpha is zero. There are two possible workarounds for this issue. The first is to use CUBLAS_PONTER_MODE_HOST instead of CUBLAS_POINTER_MODE_DEVICE, but this may require an extra device-to-host memory copy, depending on the situation. The second is to swap the (transa, alpha, A, lda) parameters with the (transb, beta, B, ldb) parameters, which would make the value pointed to by beta equal to 0.
The routine cublasCsyrk() may produce incorrect results on GPUs that implement the sm_30 architecture when the size of matrix parameter A exceeds (128M - 512) total elements.
The CUSPARSE library routines csrsv_analysis(), csrsv_solve(), csrsm_analysis(), and csrsm_solve() support the CUSPARSE_MATRIX_TYPE_GENERAL matrix type in addition to the supported matrix types already listed in the documentation.
CUDA Tools
The hardware counter (event) values may be incorrect in some cases on GPUs with compute capability (SM type) 3.5. Incorrect event values also result in incorrect metric values. These errors are more likely to occur when the same GPU is used for display and compute, or when other graphics applications are running simultaneously on the GPU.
Beginning with CUDA 5.0, the ptxas portion of the compiler generates a warning when the command line option "-abi=no" is used that indicates the option may be deprecated in a future release.
The current 5.0 linker will not support JIT to future architectures; objects will have to re-linked for each architecture.
Source-level analysis in NVIDIA Nsight Eclipse Edition and NVIDIA Visual Profiler is not available for kernels accessed through static function pointers.
Release Highlights
CUDA Dynamic Parallelism allows __global__ and __device__ functions running on the GPU to launch kernels using the familiar <<< >>> syntax and to directly call CUDA Runtime API routines (previously this ability was only available from __host__ functions).
All __device__ functions can now be separately compiled and linked using NVCC. This allows creation of closed-source static libraries of __device__ functions and the ability for these libraries to call user-defined __device__ callback functions. The linker support is considered to be a BETA feature in this release.
Nsight Eclipse Edition for Linux and Mac OS is an integrated development environment UI that allows developing, debugging, and optimizing CUDA code.
A new command-line profiler, nvprof, provides summary information about where applications spend the most time, so that optimization efforts can be properly ocused.
This release contains the following:
- NVIDIA CUDA Toolkit documentation
- NVIDIA CUDA compiler (nvcc) and supporting tools
- NVIDIA CUDA runtime libraries
- NVIDIA CUDA-GDB debugger
- NVIDIA CUDA-MEMCHECK
- NVIDIA Visual Profiler, nvprof, and command-line profiler
- NVIDIA Nsight Eclipse Edition
- NVIDIA CUBLAS, CUFFT, CUSPARSE, CURAND, Thrust, and
- NVIDIA Performance Primitives (NPP) libraries
Release notes in full:
http://docs.nvidia.com/cuda-toolkit-release-notes/index.html
Download & Instruction:
http://docs.nvidia.com/
