NVIDIA Announces CUDA Toolkit 3.2 – up to 300% Performance Increase
By
NVIDIA just sent us word that they have announced the availability of the CUDA Toolkit 3.2 production release, which provides significant performance increases, new math libraries and advanced cluster management features for developers creating next-generation GPU-accelerated applications. CUDA 3.2 can be downloaded from here if you need it.
New features and significant performance enhancements in version 3.2 include:
- Up to 300-percent performance improvement in CUDA BLAS (CUBLAS) library routines, delivering 8 times faster performance than the latest Intel MKL (Math Kernel Library)
- CUDA FFT (CUFFT) library optimizations delivering 2 – 20 times faster performance than the latest MKL
- New CURAND library for random number generation at 10-20 times faster than the latest MKL
- New CUSPARSE library of sparse matrix routines that delivers 6-30 times faster performance than the latest MKL
- A host of additional improvements to GPU debugging and performance analysis tools
Comments are closed.