Association for Computing Machinery
Coherent Shared Virtual Memory (cSVM) is highly coveted for heterogeneous architectures as it will simplify programming across different cores and manycore accelerators. In this paper, virtual L1 caches can be used to great advantage, e.g., saving energy consumption by eliminating address translation for hits. Unfortunately, multicore virtual-cache coherence is complex and costly because it requires reverse translation for any coherence request directed towards a virtual L1. The reason is the ambiguity of the virtual address due to the possibility of synonyms.