English

New Tools for Peak Memory Scheduling

Data Structures and Algorithms 2023-12-22 v1

Abstract

We study scheduling of computation graphs to minimize peak memory consumption, an increasingly critical task due to the surge in popularity of large deep-learning models. This problem corresponds to the weighted version of the classical one-shot black pebbling game. We propose the notion of a dominant schedule to capture the idea of finding the ``best'' schedule for a subgraph and introduce new tools to compute and utilize dominant schedules. Surprisingly, we show that despite the strong requirements, a dominant schedule exists for any computation graph; and, moreover, that it is possible to compute the dominant schedule efficiently whenever we can find optimal schedules efficiently for a particular class of graphs (under mild technical conditions). We apply these new tools to analyze trees and series-parallel graphs. We show that the weighted one-shot black pebbling game is strongly NP-complete even when the graph is an out-tree -- or simpler still, a pumpkin, one of the simplest series-parallel graphs. On the positive side, we design a fixed-parameter tractable algorithm to find a dominant schedule (hence also a peak memory minimizing schedule) for series-parallel graphs when parameterized by the out-degree. This algorithm runs in time 2O(dlogd)poly(n)2^{O(d \log d)} \cdot poly(n) for series-parallel graphs with nn nodes and maximum out-degree dd; for pumpkins, we can improve the dependence on dd to O(2dpoly(n))O(2^d \cdot poly(n)).

Keywords

Cite

@article{arxiv.2312.13526,
  title  = {New Tools for Peak Memory Scheduling},
  author = {Ce Jin and Manish Purohit and Zoya Svitkina and Erik Vee and Joshua R. Wang},
  journal= {arXiv preprint arXiv:2312.13526},
  year   = {2023}
}
R2 v1 2026-06-28T13:58:15.553Z