07.13 Pts, 10 minutes] Assume we have matrix of 80 by 100 and we want to cover it to index it

with a thread blocks of size 32 by 32. Analyze this case and what is the performance impact of

divergent warps?

Solution: