tag:blogger.com,1999:blog-4753335805357856016.post8863589911285146521..comments2023-07-25T14:49:41.051+06:00Comments on Unigine crew: CUDA vs OpenCL vs SPU Part IVfrustumhttp://www.blogger.com/profile/11672693079674482929noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-4753335805357856016.post-42177561114595024962011-05-30T13:38:41.031+07:002011-05-30T13:38:41.031+07:00OpenCL IS stupid. Have you tried to write a progra...OpenCL IS stupid. Have you tried to write a program in OpenCL? Storing the code into a string? Compiling OpenCL files at run-time? Managing platforms, devices, contexts, kernels, even kernel arguments!!? This is OK to build frameworks (like with the CUDA driver API), but not for regular programmers.klim8https://www.blogger.com/profile/08484133708061480569noreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-82648978273777329242011-05-30T05:20:34.518+07:002011-05-30T05:20:34.518+07:00CUDA is stupid. Programming for it will only for ~...CUDA is stupid. Programming for it will only for ~50% of the market which means you'll be doing OpenCL programming anyways.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-37681627224696500212011-01-21T09:30:03.004+06:002011-01-21T09:30:03.004+06:00i hope unigine support CUDA as well & looks li...i hope unigine support CUDA as well & looks like promisingTrololohttps://www.blogger.com/profile/13138658249858164013noreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-50738714952896280102010-09-17T22:00:50.520+07:002010-09-17T22:00:50.520+07:00It seems like command response or copying of small...It seems like command response or copying of small memory blocks is more efficient on CUDA than on OpenCL.frustumhttps://www.blogger.com/profile/11672693079674482929noreply@blogger.comtag:blogger.com,1999:blog-4753335805357856016.post-52348712957912526822010-09-17T20:38:19.714+07:002010-09-17T20:38:19.714+07:00Any idea why is the Elements 64768 benchmark has t...Any idea why is the Elements 64768 benchmark has the timings between CUDA vs OpenCL differ quite significantly? where the Elements 1036288 benchmark has much closer timings.xmanhttps://www.blogger.com/profile/05695636905017529897noreply@blogger.com