Instruction-Level Execution Migration

TL;DR: To evaluate the potential of EM 2 architectures, a series of PIN/Graite-based models of an EM 2 multicore with 64 x86 cores were developed and, under some simplifying assumptions, they were compared against corresponding directory -based cache -coherent architecture models.

Abstract: NWe introduce the Execution Migration Machine (EM 2 ), a novel data -centric multicore memory system architectu re based on computation migration. Unlike traditional distributed memory multicores, which rely on complex cache coherence protocols to move the data to the core where the computation is taking place, our scheme always moves the computation to the core whe re the data resides. By doing away with the cache coherence protocol, we are able to boost the effectiveness of per-core caches while drastically reducing hardware complexity. To evaluate the potential of EM 2 architectures, we developed a series of PIN/Gra phite-based models of an EM 2 multicore with 64 x86 cores and, under some simplifying assumptions (a timing model restricted to data memory performance, no instruction cache modeling, high -bandwidth fixed -latency interconnect allowing concurrent migrations) , compared them against corresponding directory -based cache -coherent architecture models. We justify our assumptions and show that our conclusions are valid even if our assumptions are removed. Experimental results on a range of SPLASH -2 and PARSEC benchma rks indicate that EM