02 February 2019 2 9K Report

Synchronization overheads blow up exponentially as more and more cores are deployed on a tiled mesh multicore. Synchronization costs increase as a multicore can only have a limited number of atomic operations or coherence messages running across the chip. The distance between cores also induces high latency costs. How do we circumvent these problems to accelerate performance and reduce synchronization costs?

More Masab Ahmad's questions See All
Similar questions and discussions