Four Threads, Blocked Approach

A

B

C

D

memory latency

Utilization is 41%

Previous slide

Next slide

Back to first slide

View graphic version