English

NeuMMU: Architectural Support for Efficient Address Translations in Neural Processing Units

Hardware Architecture 2019-11-19 v1 Distributed, Parallel, and Cluster Computing Machine Learning Neural and Evolutionary Computing

Abstract

To satisfy the compute and memory demands of deep neural networks, neural processing units (NPUs) are widely being utilized for accelerating deep learning algorithms. Similar to how GPUs have evolved from a slave device into a mainstream processor architecture, it is likely that NPUs will become first class citizens in this fast-evolving heterogeneous architecture space. This paper makes a case for enabling address translation in NPUs to decouple the virtual and physical memory address space. Through a careful data-driven application characterization study, we root-cause several limitations of prior GPU-centric address translation schemes and propose a memory management unit (MMU) that is tailored for NPUs. Compared to an oracular MMU design point, our proposal incurs only an average 0.06% performance overhead.

Keywords

Cite

@article{arxiv.1911.06859,
  title  = {NeuMMU: Architectural Support for Efficient Address Translations in Neural Processing Units},
  author = {Bongjoon Hyun and Youngeun Kwon and Yujeong Choi and John Kim and Minsoo Rhu},
  journal= {arXiv preprint arXiv:1911.06859},
  year   = {2019}
}
R2 v1 2026-06-23T12:17:35.492Z