CUDA 1.0 Released

NVIDIA would like to announce that version 1.0 of the CUDA Toolkit and SDK is now available.

The 1.0 release of the CUDA Toolkit includes a number of new features, including:

    Additional hardware support:

      Quadro FX 5600

      Quadro FX 4600

      GeForce 8800 Ultra

      GeForce 8600 GTS

      GeForce 8600 GT

      GeForce 8500 GT

    New operating system support (32 and 64 bit versions):

      Windows XP

      Red Hat Enterprise Linux 3.8

      Red Hat Enterprise Linux 4.3

      Red Hat Enterprise Linux 4.4

      Red Hat Enterprise Linux 5.0

      SUSE Linux Enterprise Desktop 10.0

      SUSE Linux 10.1

      SUSE Linux 10.2

    Compiler and Toolchain improvements:

      PTX ISA support

      64-bit integer support

      maxrregcount option added to NVCC

    Improved math library

      Additional functions added: sincos(), rsqrt(), exp10()

      Improved accuracy of mathematical functions

    Asynchronous launches

    Asynchronous device to device memory copy

We strongly encourage all users to upgrade to this version since it also includes many bug fixes.

The 1.0 release of the CUDA SDK includes many new code examples, including optimized versions of:

    64-bin histogram

    Separable image convolution

    Image convolution using FFT

    Sobel edge-detection filter

    The Mersenne Twister pseudo-random number generator

    Financial applications - Monte-Carlo option pricing, Black-Scholes option pricing and binomial option pricing.