01-09-2021, 12:22 PM
So in a GeForce card I get 1024 threads for the calculations.. but my GeForce card has 8000 cuda cores. Am I missing how the execution happens? As in are 8 cores needed for each thread and each calculation? It also says 128 loops in use and 1 vector.
Can someone explain how it selects the threads with cuda?
Thank you
Can someone explain how it selects the threads with cuda?
Thank you