1 paper
Tejas Shah, Bobbi Yogatama, Kyle Roarty +1
It is generally observed that the fraction of live lines in shared last-level caches (SLLC) is very small for chip multiprocessors (CMPs). This can be tackled using promotion-based…