NVIDIA CUDA 3.1 正式版发布

Edison · 发表于 2010-6-26 17:52

http://developer.nvidia.com/object/cuda_3_1_downloads.html

Edison · 发表于 2010-6-26 17:53

Release Highlights

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 Runtime 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
- Support for start/stop profiling at runtime so you can focus on critical areas of long-running applications
- Support for CUDA Driver API tracing
Math Libraries Performance Improvements, including:
- Up to 25% improvement in erfinvf() performance
- 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
- 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

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

Windows developers should be sure to check out the new debugging and profiling features in Parallel Nsight for Visual Studio at www.nvidia.com/nsight.
Please refer to the release notes and Getting Started Guides for more information.
For more information on general purpose computing features of the Fermi architecture, see: www.nvidia.com/fermi.
Note: The developer driver packages below provide baseline support for the widest number of NVIDIPro-Aducts in the smallest number of installers. More recent production driver packages for end users are available at www.nvidia.com/drivers.

[Windows XP, Windows Vista and Windows7]

[Linux]

[MacOS]

Windows XP, Windows Vista and Windows7

Developer Drivers for WinXP (257.21)	32-bit 64-bit
Developer Drivers for WinVista (257.21)	32-bit 64-bit
Notebook Developer Drivers for WinXP	32-bit 64-bit
Notebook Developer Drivers for WinVista	32-bit 64-bit
CUDA Toolkit Documentation	32-bit 64-bit	OpenCL_Implementation Notes CUDPro-Afiler Html Visual Profiler Release Notes Release Notes Getting Started Guide Windows MVIDIA CUDA C Programming Guide Reference Manual Online Reference CUDA-GDB User Manual MVIDIA CUDA C Best Practices Guide Fermi Compatibility Guide Fermi Tuning Guide License CUBLAS CUFFT CUDA Developer Guide for Optimus Platforms
GPU Computing SDK code samples	32-bit 64-bit	OpenCL Release Notes CUDA C/C++ Release Notes DirectCompute Release Notes CUDA Occupancy Calculator License
OpenCL		OpenCL Programming Guide OpenCL BestPractices Guide OpenCL JumpStart Guide OpenCL Programming Overview Visual Profiler Release Notes Compiler_Options D3D9 Sharing D3D10 Sharing D3D11 Sharing Device Attribute Query Pragma Unroll

Linux

Developer Drivers for Linux (195.36.15)	32-bit 64-bit
CUDA Toolkit for Fedora 10	32-bit 64-bit
CUDA Toolkit for RedHat Enterprise Linux 5.3	32-bit 64-bit
CUDA Toolkit for Ubuntu Linux 9.04	32-bit 64-bit
CUDA Toolkit for RedHat Enterprise Linux 4.8	32-bit 64-bit
CUDA Toolkit for OpenSUSE 11.1	32-bit 64-bit
CUDA Toolkit for SUSE Linux Enterprise Desktop 11	32-bit 64-bit
CUDA Toolkit Documentation		OpenCL_Implementation Notes CUDPro-Afiler Html Visual Profiler Release Notes Release Notes Getting Started Guide CUDA C Programming Guide Reference Manual Online Reference CUDA-GDB User Manual CUDA C Best Best Practices Guide Fermi Compatibility Guide Fermi Tuning Guide DirectCompute Release Notes License CUBLAS CUFFT CUDA Developer Guide for Optimus Platforms
GPU Computing SDK code samples and more	download	CUDA C/C++ Release Notes CUDA Occupancy Calculator License
OpenCL		OpenCL Programming Guide OpenCL BestPractices Guide OpenCL JumpStart Guide OpenCL Programming Overview Visual Profiler Release Notes Compiler_Options D3D9 Sharing D3D10 Sharing D3D11 Sharing Device Attribute Query Pragma Unroll

MacOS

Developer Drivers for MacOS	download
CUDA Toolkit	download	CUDPro-Afiler Html Visual Profiler Release Notes Release Notes Getting Started Guide CUDA C Programming Guide Reference Manual Online Reference License CUBLAS CUFFT
GPU Computing SDK code samples and more	download	CUDA C/C++ Release Notes CUDA Occupancy Calculator License

snock · 发表于 2010-6-26 19:10

爽啊，能用了

snock · 发表于 2010-6-26 19:11

本帖最后由 snock 于 2010-6-26 19:17 编辑

前一段正好在GTX260+上做自适应滤波，觉得效率太低了，就等Fermi的专用驱动了

gz_easy · 发表于 2010-6-26 19:16

CUDA持续升级，可作为CUDA应用的badaboom却要到今年第四季度才支持GTX400系列。

66666 · 发表于 2010-6-26 19:18

badaboom估计在等NV新的CUDA H264编码吧

zhu198352 · 发表于 2010-6-26 19:20

感觉没啥实际用处

disruptor · 发表于 2010-6-26 23:53

CUDA的函数库真是越来越强大了

PaulWong · 发表于 2010-6-27 00:33

自己用不上就是没用？我用A卡也用不了，不过也不敢说没用。

qingren488 · 发表于 2010-6-27 14:11

发布了..我的96能支持么？？

feiying2222 · 发表于 2010-6-27 16:27

本帖最后由 feiying2222 于 2010-6-27 16:33 编辑

Release HighlightsGPUDirect(tm) gives 3rd party devices direct access to CUDA MemorySupport ...
Edison 发表于 2010-6-26 17:53

呵呵，正在用CUDA 3.0编东西，这个消息对我很重要，谢谢！

feiying2222 · 发表于 2010-6-27 16:34

Edison 发表于 2010-6-26 17:52

另外，斑竹最近有没有看到GEFORCE系列破解双精度能力方面的消息？一直等待强人破解，NV对双精度限制得太厉害了，但实际上计算方面双精度是最常用的。

feiying2222 · 发表于 2010-6-27 17:16

所以等待超级大牛人出现啊。。。

ga1axy · 发表于 2010-6-27 17:36

CUDA的函数库真是越来越强大了

帐号		自动登录	找回密码
密码			注册