Lune

MICRO2022Top-tier venue

Berti: an Accurate Local-Delta Data Prefetcher

Agustín Navarro-Torres, Biswabandan Panda, Jesús Alastruey-Benedé, Pablo Ibáñez, Víctor Viñals Yúfera, Alberto Ros

Abstract

Data prefetching is a technique that plays a crucial role in modern high-performance processors by hiding long latency memory accesses. Several state-of-the-art hardware prefetchers exploit the concept of deltas, defined as the difference between the cache line addresses of two demand accesses. Existing delta prefetchers, such as best offset prefetching (BOP) and multi-lookahead prefetching (MLOP), train and predict future accesses based on global deltas. We observed that the use of global deltas results in missed opportunities to anticipate memory accesses.

In this paper, we propose Berti, a first-level data cache prefetcher that selects the best local deltas, i.e., those that consider only demand accesses issued by the same instruction. Thanks to a high-confidence mechanism that precisely detects the timely local deltas with high coverage, Berti generates accurate prefetch requests. Then, it orchestrates the prefetch requests to the memory hierarchy, using the selected deltas.

Our empirical results using ChampSim and SPEC CPU2017 and GAP workloads show that, with a storage overhead of just 2.55 KB, Berti improves performance by 8.5% compared to a baseline IP-stride and 3.5% compared to IPCP, a state-of-theart prefetcher. Our evaluation also shows that Berti reduces dynamic energy at the memory hierarchy by 33.6% compared to IPCP, thanks to its high prefetch accuracy.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

21 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Pythia: A Customizable Hardware Prefetching Framework Using Online Reinforcement Learning cites A hierarchical neural model of data prefetching
  • Pythia: A Customizable Hardware Prefetching Framework Using Online Reinforcement Learning cites Bouquet of Instruction Pointers: Instruction Pointer Classifier-based Spatial Hardware Prefetching
  • To Cross, or Not to Cross Pages for Prefetching? cites Pythia: A Customizable Hardware Prefetching Framework Using Online Reinforcement Learning
  • To Cross, or Not to Cross Pages for Prefetching? cites A hierarchical neural model of data prefetching
  • To Cross, or Not to Cross Pages for Prefetching? cites Bouquet of Instruction Pointers: Instruction Pointer Classifier-based Spatial Hardware Prefetching
  • A New Formulation of Neural Data Prefetching cites Pythia: A Customizable Hardware Prefetching Framework Using Online Reinforcement Learning
  • A New Formulation of Neural Data Prefetching cites A hierarchical neural model of data prefetching
  • A New Formulation of Neural Data Prefetching cites Bouquet of Instruction Pointers: Instruction Pointer Classifier-based Spatial Hardware Prefetching
  • Decoupled Vector Runahead cites Pythia: A Customizable Hardware Prefetching Framework Using Online Reinforcement Learning
  • Decoupled Vector Runahead cites A hierarchical neural model of data prefetching
  • Decoupled Vector Runahead cites Bouquet of Instruction Pointers: Instruction Pointer Classifier-based Spatial Hardware Prefetching
  • Scalar Vector Runahead cites Pythia: A Customizable Hardware Prefetching Framework Using Online Reinforcement Learning
  • Scalar Vector Runahead cites Bouquet of Instruction Pointers: Instruction Pointer Classifier-based Spatial Hardware Prefetching
  • Scalar Vector Runahead cites Decoupled Vector Runahead
  • A Two Level Neural Approach Combining Off-Chip Prediction with Adaptive Prefetch Filtering cites Bouquet of Instruction Pointers: Instruction Pointer Classifier-based Spatial Hardware Prefetching
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers21

  1. A Two Level Neural Approach Combining Off-Chip Prediction with Adaptive Prefetch FilteringAlexandre Valentin Jamet, Georgios Vavouliotis, Daniel A. Jiménez, Lluc Alvarez et al.HPCA 2024 · 18 citations
  2. A New Formulation of Neural Data PrefetchingQuang Duong, Akanksha Jain, Calvin LinISCA 2024 · 16 citations
  3. Decoupled Vector RunaheadAjeya Naithani, Jaime Roelandts, Sam Ainsworth, Timothy M. Jones et al.MICRO 2023 · 15 citations
  4. Scalar Vector RunaheadJaime Roelandts, Ajeya Naithani, Sam Ainsworth, Timothy M. Jones et al.MICRO 2024 · 11 citations
  5. To Cross, or Not to Cross Pages for Prefetching?Georgios Vavouliotis, Martí Torrents, Boris Grot, Kleovoulos Kalaitzidis et al.HPCA 2025 · 10 citations

Builds on4

  1. A hierarchical neural model of data prefetchingZhan Shi, Akanksha Jain, Kevin Swersky, Milad Hashemi et al.ASPLOS 2021 · 100 citations
  2. Bouquet of Instruction Pointers: Instruction Pointer Classifier-based Spatial Hardware PrefetchingSamuel Pakalapati, Biswabandan PandaISCA 2020 · 97 citations
  3. Pythia: A Customizable Hardware Prefetching Framework Using Online Reinforcement LearningRahul Bera, Konstantinos Kanellopoulos, Anant Nori, Taha Shahroodi et al.MICRO 2021 · 95 citations
  4. A Cost-Effective Entangling Prefetcher for InstructionsAlberto Ros, Alexandra JimboreanISCA 2021 · 31 citations

Related papers

  1. CLIP: Load Criticality based Data Prefetching for Bandwidth-constrained Many-core SystemsBiswabandan PandaMICRO 2023 · 21 citations
  2. Reducing Load Latency with Cache Level PredictionMajid Jalili, Mattan ErezHPCA 2022 · 17 citations
  3. Integrating Prefetcher Selection with Dynamic Request Allocation Improves Prefetching EfficiencyMengming Li, Qijun Zhang, Yongqing Ren, Zhiyao XieHPCA 2025 · 6 citations
  4. Twig: Profile-Guided BTB Prefetching for Data Center ApplicationsTanvir Ahmed Khan, Nathan Brown, Akshitha Sriraman, Niranjan K. Soundararajan et al.MICRO 2021 · 33 citations
  5. ICP: Exploiting Instruction Correlation for Prefetching Irregular Memory AccessesMengming Li, Chenlu Miao, Buqing Xu, Qijun Zhang et al.ISCA 2026
Dusk over the sea between two cliffs drawn in fine vertical lines