XBOF: A Cost-Efficient CXL JBOF with Inter-SSD Compute Resource Sharing

2025-09-12

AI Summary

Enterprise SSDs integrate substantial compute resources to handle I/O bursts, yet under JBOF architectures, these resources suffer from low utilization, inflating per-unit storage cost. This paper proposes XBOF, the first architecture enabling decoupled and decentralized cross-SSD sharing of compute capabilities. Leveraging the CXL cache-coherence protocol, XBOF dynamically schedules idle SSDs’ compute resources to serve overloaded SSDs, incorporating modular function partitioning, distributed resource scheduling, and cross-device metadata acceleration. Experiments show that XBOF improves SSD compute-resource utilization by 50.4% and reduces overall storage-system cost by 19.0% versus conventional JBOF, with negligible performance overhead. Its core contribution is breaking the SSD-level resource isolation barrier, establishing the first cross-SSD compute-sharing architecture supporting fine-grained allocation, low runtime overhead, and hardware-enforced cache coherence.

Technology Category

• Search and Optimization: Distributed Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Machine Learning: Hardware-aware ML

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Decentralized Web and Fediverse systems
• Economics, Online Markets and Human Computation: The sharing economy
• Social Networks and Social Media: Computational social science

Abstract

Enterprise SSDs integrate numerous computing resources (e.g., ARM processor and onboard DRAM) to satisfy the ever-increasing performance requirements of I/O bursts. While these resources substantially elevate the monetary costs of SSDs, the sporadic nature of I/O bursts causes severe SSD resource underutilization in just a bunch of flash (JBOF) level. Tackling this challenge, we propose XBOF, a cost-efficient JBOF design, which only reserves moderate computing resources in SSDs at low monetary cost, while achieving demanded I/O performance through efficient inter-SSD resource sharing. Specifically, XBOF first disaggregates SSD architecture into multiple disjoint parts based on their functionality, enabling fine-grained SSD internal resource management. XBOF then employs a decentralized scheme to manage these disaggregated resources and harvests the computing resources of idle SSDs to assist busy SSDs in handling I/O bursts. This idea is facilitated by the cache-coherent capability of Compute eXpress Link (CXL), with which the busy SSDs can directly utilize the harvested computing resources to accelerate metadata processing. The evaluation results show that XBOF improves SSD resource utilization by 50.4% and saves 19.0% monetary costs with a negligible performance loss, compared to existing JBOF designs.

Problem

• Addressing SSD resource underutilization in JBOF systems
• Enabling efficient inter-SSD compute resource sharing
• Reducing monetary costs while maintaining I/O performance

Innovation

• Inter-SSD compute resource sharing
• Disaggregated SSD architecture management
• CXL-enabled decentralized resource utilization