Mole: Tier-Specific Hotness Profiling Driven Memory Tiering for Multi-Tiered Memory Systems

2026-10-08

AI Summary

This study addresses capacity contention and migration failures in the fast tier of hierarchical memory systems caused by greedy promotion and stepwise demotion. To mitigate these issues, this work proposes a staging demotion mechanism that exploits hot-cold asymmetry. By decoupling promotion from demotion objectives, cold pages are directly migrated to the bottom tier to free fast-tier capacity. Furthermore, a low-overhead, tier-differentiated heat profiling technique is designed to accurately capture page state transitions and guide objective-oriented promotion decisions. Experimental results demonstrate that the proposed approach significantly reduces migration failure rates under dynamic access patterns while effectively improving the overall performance and resource utilization of multi-tiered memory systems.

Abstract

Multi-tiered memory systems combine fast, small upper tiers with slow, large lower tiers to improve performance and cost efficiency. Existing designs, such as AutoTiering and MTM, rely on greedy promotion and stepwise demotion, which can intensify contention for limited capacity in faster tiers. We observe that promotions directly improve performance, whereas demotions primarily reclaim space. Based on this asymmetry, we propose Mole, a memory tiering system that separates promotion and demotion destinations. Mole employs staging demotion to move cold pages directly to the lowest tier, bypassing intermediate tiers, and targeted promotion to place hot pages in tiers that match their current hotness. The lowest tier serves as a staging area from which pages can be promoted when they become hot again. This separation preserves intermediate-tier capacity for promotions, reducing tier contention and migration failures. However, staging demotion requires timely identification of pages that become hot after demotion. To meet this requirement with low profiling overhead, Mole uses tier-specific profiling: it profiles the lowest tier at high frequency to detect reactivated hot pages, while profiling upper tiers at lower frequency to identify cold pages. Experimental results show that Mole reduces migration failures and improves performance under dynamic access patterns.

Problem

• Multi-tiered memory systems
• Memory tiering
• Tier contention
• Page migration
• Hotness profiling

Innovation

• Memory Tiering
• Staging Demotion
• Targeted Promotion
• Tier-Specific Profiling
• Multi-Tiered Memory Systems