Reproduce “Memory Fit: Architectural Support for Page Placement by Learning Models of Good Behavior” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Chun Deng, Hui Sub Shim, Evan Laufer, Anna Eaton, David Tennenhouse, Tsachy Weissman, Philip Levis
Memory Fit uses learned scores to help Linux place pages across DRAM and PCM while leaving placement policy to the operating system. The scores reflect how reads, writes, and their ordering fit memory rather than access frequency alone. In a Redis placement replay, it recovered 80.0% of the all-PCM-to-all-DRAM cycle gap, versus 18.7% for TPP; the replay isolates placement quality, not migration costs.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.