CXL Topology-Aware and Expander-Driven Prefetching: Unlocking SSD Performance

TL;DR

This paper introduces ExPAND, a novel CXL-SSD prefetcher that improves performance by offloading cache prefetching, using a heterogeneous prediction algorithm, and providing accurate latency estimations, significantly boosting graph and CPU workloads.

Contribution

ExPAND is the first CXL-SSD prefetcher that leverages topology awareness and expander-driven algorithms to enhance prefetching accuracy and performance.

Findings

ExPAND improves graph application performance by 9.0×.

ExPAND enhances SPEC CPU performance by 14.7×.

It surpasses existing CXL-SSD prefetching strategies.

Abstract

Integrating compute express link (CXL) with SSDs allows scalable access to large memory but has slower speeds than DRAMs. We present ExPAND, an expander-driven CXL prefetcher that offloads last-level cache (LLC) prefetching from host CPU to CXL-SSDs. ExPAND uses a heterogeneous prediction algorithm for prefetching and ensures data consistency with CXL.mem's back-invalidation. We examine prefetch timeliness for accurate latency estimation. ExPAND, being aware of CXL multi-tiered switching, provides end-to-end latency for each CXL-SSD and precise prefetch timeliness estimations. Our method reduces CXL-SSD reliance and enables direct host cache access for most data. ExPAND enhances graph application performance and SPEC CPU's performance by 9.0$\times$ and 14.7$\times$, respectively, surpassing CXL-SSD pools with diverse prefetching strategies.