{"id":2520,"date":"2016-05-28T09:13:19","date_gmt":"2016-05-28T13:13:19","guid":{"rendered":"http:\/\/local.brightwhiz\/?p=2520"},"modified":"2016-05-28T09:13:19","modified_gmt":"2016-05-28T13:13:19","slug":"cuda-8-toolkit-release-candidate","status":"publish","type":"post","link":"http:\/\/local.brightwhiz\/cuda-8-toolkit-release-candidate\/","title":{"rendered":"The New CUDA 8 Toolkit Release Candidate Now Available"},"content":{"rendered":"
NVIDIA has announced the immediate availability of CUDA<\/a> 8 Toolkit Release Candidate. The toolkit can be downloaded for free by the members of the ComputeWorks Developer Program.<\/p>\n <\/p>\n CUDA is a set of parallel computing Application Programming Interfaces (APIs) developed by NVIDIA<\/a>. CUDA enables developers to use the toolkit to program on compatible graphics processing units (GPU<\/a>) for general purpose applications (GPGPU).<\/p>\n The new CUDA 8 Toolkit Release Candidate now comes with full support for all of NVIDIAs Pascal architecture based GPUs. Developers now get simplified Unified Memory programming. This includes support for large datasets and concurrent data on Pascal GPUs.<\/p>\n For easy performance optimization, there is automated critical path analysis available to CUDA 8 users and more. You can get more details about this new toolkit by visiting the CUDA 8 Toolkit Release Candidate page here. You can also download the toolkit<\/a> while you are on that page.<\/p>\n","protected":false},"excerpt":{"rendered":" NVIDIA has announced the immediate availability of CUDA 8 Toolkit Release Candidate. The toolkit can be downloaded for free by the members of the ComputeWorks Developer Program.<\/p>\n","protected":false},"author":1,"featured_media":2521,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[13,28,16,18],"tags":[33,111,143,146,172,184,220,262,280,291,350,412,433,449,471,544,545,591],"yoast_head":"\nCUDA 8 Toolkit Release Candidate Highlights<\/h2>\n