Download

Fixes:
- GPUDirect(tm) gives 3rd party devices direct access to CUDA Memory
- Support for 16-way concurrency allows up to 16 different kernels to run at the same time on Fermi architecture GPUs
- Runtime / Driver interoperability enables applications to mix-n-match use of the CUDA Driver API with CUDA C Runtim and math libraries via buffer sharing and context migration

New language features added to CUDA C / C++ include:
        - Support for printf() in device code
        - Support for function pointers and recursion make it easier to port many existing algorithms to Fermi GPUs
- Unified Visual Profiler now supports both CUDA C/C++ and OpenCL, and now includes support for CUDA Driver API tracing

Math Libraries Performance Improvements, including:
        - Improved performance of selected transcendental functions from the log, pow, erf, and gamma families
        - Significant improvements in double-precision FFT performance on Fermi-architecture GPUs for 2^n transform sizes
        - Streaming API now supported in CUBLAS for overlapping copy and compute operations
        - CUFFT Real-to-complex (R2C) and complex-to-real (C2R) optimizations for 2^n data sizes
        - Improved performance for GEMV and SYMV subroutines in CUBLAS
        - Optimized double-precision implementations of divide and reciprocal routines for the Fermi architecture

New and updated SDK code samples demonstrating how to use:
        - Function pointers in CUDA C/C++ kernels
        - OpenCL / Direct3D buffer sharing
        - Hidden Markov Model in OpenCL
        - Microsoft Excel GPGPU example showing how to run an Excel function on the GPU