Pangloss: a novel Markov chain prefetcher

TL;DR

Pangloss is a high-performance data prefetcher that uses an efficient Markov chain approximation to predict complex memory access patterns, significantly improving cache speedups in multi-core systems.

Contribution

It introduces a novel Markov chain-based prefetcher with a compact representation capable of reconstructing complex access patterns and obfuscated transitions.

Findings

Achieves 1.7% and 3.2% speedup over state-of-the-art baselines.

Combined L1 and L2 prefetchers yield up to 8.4% speedup.

Shows notable performance improvements in multi-core evaluations.

Abstract

We present Pangloss, an efficient high-performance data prefetcher that approximates a Markov chain on delta transitions. With a limited information scope and space/logic complexity, it is able to reconstruct a variety of both simple and complex access patterns. This is achieved by a highly-efficient representation of the Markov chain to provide accurate values for transition probabilities. In addition, we have added a mechanism to reconstruct delta transitions originally obfuscated by the out-of-order execution or page transitions, such as when streaming data from multiple sources. Our single-level (L2) prefetcher achieves a geometric speedup of 1.7% and 3.2% over selected state-of-the-art baselines (KPCP and BOP). When combined with an equivalent for the L1 cache (L1 & L2), the speedups rise to 6.8% and 8.4%, and 40.4% over non-prefetch. In the multi-core evaluation, there seems to be a considerable performance improvement as well.