tag:blogger.com,1999:blog-4753335805357856016.post2551044692797363726..comments2023-07-25T14:49:41.051+06:00Comments on Unigine crew: CUDA vs OpenCL vs DirectComputefrustumhttp://www.blogger.com/profile/11672693079674482929noreply@blogger.comBlogger7125tag:blogger.com,1999:blog-4753335805357856016.post-4136121760887808242013-03-28T08:10:06.855+07:002013-03-28T08:10:06.855+07:00Could you share the source code of these projectsCould you share the source code of these projectspohsibrutrahttps://www.blogger.com/profile/12307158619847806999noreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-16539177744586834112010-11-17T08:21:58.290+06:002010-11-17T08:21:58.290+06:00Can I please have the algorithm or code for the qu...Can I please have the algorithm or code for the quicksort in cuda? I've to compare the performance of quicksort in various gpus in my lab.<br />Thanks in advance.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-40647621036605522392010-04-11T07:26:21.651+07:002010-04-11T07:26:21.651+07:00which API do you like working in most?which API do you like working in most?EMCPhttps://www.blogger.com/profile/02718535225865223761noreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-8864123540638725472010-02-04T15:24:06.252+06:002010-02-04T15:24:06.252+06:00The result is: "Display driver stopped respon...The result is: "Display driver stopped responding and has recovered".frustumhttps://www.blogger.com/profile/11672693079674482929noreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-57945052977759698262010-02-04T15:03:31.063+06:002010-02-04T15:03:31.063+06:00What are the OpenCL-results on the ATI-cards?What are the OpenCL-results on the ATI-cards?Unknownhttps://www.blogger.com/profile/15235610158997955895noreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-51152115357690790012010-02-03T10:14:46.197+06:002010-02-03T10:14:46.197+06:00This implementation uses shared memory and same ke...This implementation uses shared memory and same kernels for all API. Bitonic block size is 512, transposition block size is 16x16.frustumhttps://www.blogger.com/profile/11672693079674482929noreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-62286946196683511872010-02-02T21:48:54.536+06:002010-02-02T21:48:54.536+06:00are u using shared memory in your implementation?
...are u using shared memory in your implementation?<br />must say the timings look very odd ... the 4850 is the one with the worst specs for these kind of computations for sure the 5850 is much much faster ...Anonymousnoreply@blogger.com