GUMPIM is proposed, a PIM system that allows interleaved CPU pages and non-interleaved PIM pages to coexist in a Unitary and Malleable memory space with Guaranteed PIM page allocation and enables zero-copy during PIM task offloading and maintains CPU memory bandwidth while ensuring low PIM offloading overhead.
Yilong Zhao, Fang-Xin Liu, Yi-Wei Hu et al.· ACM Transactions on Architec...· 2 citations
This work proposes CURE, a budget-aware dynamic repair tree designed to repair errors at uncertainty focal points without incurring prohibitive tree-verification overheads, and provides a plug-and-play repair module compatible with standard parallel drafting frameworks.
Ao-Fan Liu, Jing Meng, Fangxin Liu et al.· 1 citation
PCoMoE is presented, a path-compositional execution framework that shifts MoE inference from coarse-grained expert selection to fine-grained path composition and achieves up to a 1.31x end-to-end inference speedup while enhancing model accuracy by 10%.
Zi-Yan Gan, Fangxin Liu, Chenyang Guan et al.· 0 citations
