Principles of Memory Hierarchy Optimization 2021

Workshop Program

PPoPP Workshop on Principles of Memory Hierarchy Optimization (PMHO)
Sunday 2/28/2021 9am to 1pm US EST (China 10pm to 2am, Europe 3pm to 7pm, US PST 6am to 10pm, US MST 7 am to 11 am)
Opening and Logistics (9am)
Session 1: Algorithm Locality (9:10am to 9:50am)
Data Movement Distance: An Asymptotic and Algorithmic Measure of Memory Cost, Chen Ding, with Donovan Snyder, University of Rochester, unpublished (video)
Processor-Aware Cache-Oblivious Algorithms, Yuan Tang and Weiguo Gao, Fudan University, arxiv 2020
Keynote (10am to 10:30am)
KEYNOTE: Working Sets, Locality, and System Performance, Peter Denning, Naval Post Graduate School, ACM Computing Surveys 2021 (video)
Session 2: Optimality (10:30am to 11:30am)
Some mathematical facts about optimal cache replacement, Pierre Michaud, Inria, ACM TACO 2016 (video)
Practical Bounds on Optimal Caching with Variable Object Sizes, Nathan Beckmann, Cargenie Mellon University, ACM SIGMetrics 2018 (video)
Category Theory in the Memory Hierarchy, Wesley Smith, University of Rochester, Unpublished (video)
Session 3: Applied Theories (11:40am to 1pm)
Optimal Data Placement for Heterogeneous Cache, Memory, and Storage Systems, Lei Zhang and Ymir Vigfusson, with Reza Karimi and Irfan Ahmad, Emory University, ACM SIGMetrics 2020 (video)
PHOEBE: Reuse-Aware Online Caching with Reinforcement Learning for Emergin Storage Models, Pengcheng Li, with Nan Wu, Alibaba, arxiv 2020 (video)
Data-Model Convergence and Runtime Support for Data Analytics and Graph Algorithms, Antonino Tumeo, Pacific Northwest National Laboratory, ACM Computing Frontiers 2019 (video)
Performance Prediction Toolkit for GPUs Cache Memories, Yehia Arafa and Hameed Badawy, New Mexico State University, ACM International Conference on Supercomputing (ICS) 2020 (video)

Call for Presentations

Data movement is now often the most significant constraint on speed, throughput, and power consumption for applications in a wide range of domains. Unfortunately, modern memory systems are almost always hierarchical, parallel, dynamically allocated and shared, making them hard to model and analyze. In addition, the complexity of these systems is rapidly increasing due to the advent of new technologies such as Intel Optane and high-bandwidth memory (HBM).

Conventional solutions to memory hierarchy optimization problems are often compartmentalized and heuristic-driven; in the modern era of complex memory systems these lack the rigor and robustness to be fully effective. Going forward, research on performance in the memory hierarchy must adapt, ideally creating theories of memory that aim at formal, rigorous performance and correctness models, as well as optimizations that are based on mathematics, ensure reproducible results, and have provable guarantees.

PMHO 2021 is a forum dedicated to the theoretical aspect of memory hierarchies as well as their programming models and optimization.

Format and Topics

PMHO 2021 will be a specialized and topic-driven workshop to present ongoing, under review, or previously published work related to the following non-exhaustive topic list:

  • Mathematical and theoretical models of memory hierarchies including CPU/GPU/accelerator caches, software caching systems, key-value stores, network caches and storage system caches
  • Algorithmic and formal aspects of cache management including virtual memory
  • Programming models and compiler optimization of hierarchical complex memory systems

There will be no formal peer review process or publication, and presentations will be a mix of selections from submissions and invited presentations.

The workshop will take place online Sunday February 28, 2021.


Submit your presentation proposal to Submissions should consist of an one-page abstract of the topic you intend to present alongside any (optional) pertinent publications or preprints.

The submission deadline is Friday January 22, 2021 AOE.



  • Chen Ding, University of Rochester
  • Xipeng Shen, North Carolina State University


  • Wesley Smith, University of Rochester