14 DAY TRIAL // NVIDIA has released the CUDA 2.2 toolkit and SDK, which allows applications coded in C to use not just the CPU, but also the GPU of a machine, enhancing the overall processing speed. Among the other noteworthy additions are the ability to zero-copy, or to read and write directly from the pinned system memory, and the ability to have multiple GPUs access the same system memory information. OS X is still supported, with the Windows version of the toolkit and SDK adding support for Windows 7.
“NVIDIA announced today it has released version 2.2 of the CUDA Toolkit and SDK for GPU Computing,” the announcement on NVIDIA's website reads. “This latest release supports several significant new features that deliver a major leap forward in getting the most performance out of NVIDIA’s massively parallel CUDA-enabled GPUs. In addition, version 2.2 of the CUDA Toolkit includes support for Windows 7, the upcoming OS from Microsoft that embraces GPU Computing,” the company specializing in the development of graphics processing units and chipset technologies says.
CUDA is the only C language environment that enables programmers and developers to write software to solve complex computational problems in a fraction of the time, by tapping into the many-core parallel processing power of GPUs. According to the release notes, additional new features in CUDA Toolkit 2.2 include: Visual Profiler for the GPU; Improved OpenGL Interop; Texture from Pitch Linear Memory; Zero-copy; Pinned Shared Sysmem; Asynchronous memcopy on Vista; Hardware Debugger for the GPU and Exclusive Device Mode.
New Mac OS X-specific features comprise: Pinned Memory Support; Function attribute query; 2D Texture reads from pitch linear memory; Flags for event creation; New device management and context creation flags; Improved runtime device management; Driver/runtime version query functions and New device attribute queries. For a list of known issues, see the full release notes here.
Developers can download the latest CUDA Toolkit, SDK, and drivers now using the link below.