Quote:The second reason is that it would be hard to sync the stuff on the GPU and the stuff on the CPU, much in the same way that it would be hard to sync everything if an extra core was used, as Dolphin would have to decide which instructions to send to which things, and carry the results to the right place when they were needed, hence a performance decrease (You're unable to multithread code written for one thread).
This part makes little sense. CUDA threads are not like windows or posix threads, I'll just leave it at that.
Quote:I think I should have left this at 'No'.
Yes, I think you should have.
"Normally if given a choice between doing something and nothing, I’d choose to do nothing. But I would do something if it helps someone else do nothing. I’d work all night if it meant nothing got done."
-Ron Swanson
"I shall be a good politician, even if it kills me. Or if it kills anyone else for that matter. "
-Mark Antony
-Ron Swanson
"I shall be a good politician, even if it kills me. Or if it kills anyone else for that matter. "
-Mark Antony