Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 10, 2025Open Access

HybridTier: an Adaptive and Lightweight CXL-Memory Tiering System

View Full Paper
Ask AI
Bookmark
Share

Authors

KSKai-Tai SongJYJiacheng YangZWZixuan Wang

Discussion

Loading...

Member takes

Overview

Proposing HybridTier, an adaptive memory tiering system that optimizes data hotness tracking while reducing cache overhead.

Key Points

  • HybridTier reduces memory overhead by tracking data accesses probabilistically, enhancing efficiency with minor trade-offs.
  • The system can achieve up to 91% performance improvement over prior tiering systems with less memory overhead.
  • Using lightweight data structures, HybridTier optimizes cache accesses, recording data hotness with reduced misses.
  • It addresses the challenges of adapting to data access distributions while minimizing cache overhead from tiering.

Cite This Study

Song et al. (2025) studied this question.

synapsesocial.com/papers/68c1b81f54b1d3bfb60ec60ehttps://doi.org/10.1145/3676642.3736119
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 2 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1DaeMon: Architectural Support for Efficient Data Movement in Fully Disaggregated Systems2023 · 14 citations
  2. 2HeMem2021 · 109 citations