Author Topic: CUDA Toolkit 3.0 Beta available to GPU Computing registered developers  (Read 3761 times)

0 Members and 1 Guest are viewing this topic.


  • Global Moderator
  • Hero Member
  • *****
  • Posts: 4189
    • View Profile
The CUDA Toolkit 3.0 Beta is now available to GPU Computing registered developers.

Highlights for this release include:

    * CUDA Driver / Runtime Buffer Interoperability, which allows applications using the CUDA Driver API to also use libraries implemented using the CUDA C Runtime.
    * A new, separate version of the CUDA C Runtime (CUDART) for debugging in emulation-mode.
    * C++ Class Inheritance and Template Inheritance support for increased programmer productivity
    * A new unified interoperability API for Direct3D and OpenGL, with support for:
          o OpenGL texture interop
          o Direct3D 11 interop support
    * cuda-gdb hardware debugging support for applications that use the CUDA Driver API
    * New CUDA Memory Checker reports misalignment and out of bounds errors, available as a debugging mode within cuda-gdb and also as a stand-alone utility.
    * CUDA Toolkit libraries are now versioned, enabling applications to require a specific version, support multiple versions explicitly, etc.
    * CUDA C/C++ kernels are now compiled to standard ELF format
    * Support for all the OpenCL features in the latest R195.39 beta driver:
          o Double Precision
          o OpenGL Interoperability, for interactive high performance visualization
          o Query for Compute Capability, so you can target optimizations for GPU architectures (cl_nv_device_attribute_query)
          o Ability to control compiler optimization settings, etc. via support for NVIDIA Compiler Flags (cl_nv_compiler_options)
          o OpenCL Images support, for better/faster image filtering
          o 32-bit Atomics for fast, convenient data manipulation
          o Byte Addressable Stores, for faster video/image processing and compression algorithms
          o Support for the latest OpenCL spec revision 48 and latest official Khronos OpenCL headers as of 11/1/2009
    * Early support for the Fermi architecture, including:
          o Native 64-bit GPU support
          o Multiple Copy Engine support
          o ECC reporting
          o Concurrent Kernel Execution
          o Fermi HW debugging support in cuda-gdb

For more information on general purpose computing features of the Fermi architecture, see:

Windows developers should be sure to sign up for the Nexus (codename) beta program, and test drive the integrated support for GPU hardware debugging, profiling, and platform trace/analysis features at: