The Dawn of Disaggregation and the Coherence Conundrum: A Call for Federated Coherence

2025-04-22

AI Summary

To address scalability bottlenecks and hardware overheads caused by global cache coherence in disaggregated memory data centers, this paper proposes Federated Consistency—a novel coherence model that enforces strong consistency only within individual nodes while abandoning cross-node global coherence. It formally defines node-level local consistency semantics for the first time, departing from conventional system-wide or hierarchical coherence paradigms and enabling direct reuse of existing CPU core protocols (e.g., MESI). Integrated with a lightweight synchronization library and a disaggregation-aware programming model, the approach significantly reduces interconnect bandwidth pressure and hardware complexity. Experimental evaluation demonstrates that, while preserving application correctness, Federated Consistency improves system scalability and deployment feasibility. It establishes a new coherence foundation for high-performance disaggregated memory systems—striking a balance between conceptual simplicity and practical applicability.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Multiagent Systems: Coordination and Collaboration

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Security and Privacy: Data transparency and provenance

Abstract

Disaggregated memory is an upcoming data center technology that will allow nodes (servers) to share data efficiently. Sharing data creates a debate on the level of cache coherence the system should provide. While current proposals aim to provide coherence for all or parts of the disaggregated memory, we argue that this approach is problematic, because of scalability limitations and hardware complexity. Instead, we propose and formally define federated coherence, a model that provides coherence only within nodes, not across nodes. Federated coherence can use current intra-node coherence provided by processors without requiring expensive mechanisms for inter-node coherence. Developers can use federated coherence with a few simple programming paradigms and a synchronization library. We sketch some potential applications.

Problem

• Addressing scalability issues in disaggregated memory coherence
• Reducing hardware complexity in shared data systems
• Proposing federated coherence for intra-node efficiency

Innovation

• Proposes federated coherence model
• Uses intra-node coherence only
• Simplifies programming with synchronization library