EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

The Landscape of GPU-Centric Communication

Didem Unat, Ilyas Turimbetov, Mohammed Issa, Dogan Sagbili, Flavio Vella, Daniele Sensi, Ismayil Ismayilov

ACM Computing Surveys · 30.09.2026 · 58 (12) · 1-36

DOI: 10.1145/3813799 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Computing Surveys, online erschienen am 09.06.2026. Verfasst von Didem Unat und 6 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 165 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: European Research Council, European Union’s Horizon Europe, Sapienza University.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In recent years, GPUs have become the preferred accelerators for HPC and ML applications due to their parallelism and high memory bandwidth. While GPUs boost computation, inter-GPU communication can create scalability bottlenecks, especially as the number of GPUs per node and cluster grows. Traditionally, the CPU managed multi-GPU communication, but advancements in GPU-centric communication now challenge this CPU dominance by reducing its involvement, granting GPUs more autonomy in communication tasks, and addressing mismatches in multi-GPU communication and computation. This article provides a landscape of GPU-centric communication, focusing on vendor mechanisms and user-level library supports. It aims to clarify the complexities and diverse options in this field, define the terminology, and categorize existing approaches within and across nodes. The article discusses vendor-provided mechanisms for communication and memory management in multi-GPU execution and reviews major communication libraries, their benefits, challenges, and performance insights. Then, it explores key research paradigms, future outlooks, and open research questions. By extensively describing GPU-centric communication techniques across the software and hardware stacks, we provide researchers, programmers, engineers, and library designers insights on how to exploit multi-GPU systems at their best.

Förderung

  • European Research Council
  • European Union’s Horizon Europe
  • Sapienza University

Zitieren

APA

Unat, D., Turimbetov, I., Issa, M., Sagbili, D., Vella, F., Sensi, D., & Ismayilov, I. (2026). The Landscape of GPU-Centric Communication. ACM Computing Surveys, 58(12), 1-36. https://doi.org/10.1145/3813799

Vancouver

Unat D, Turimbetov I, Issa M, Sagbili D, Vella F, Sensi D, et al. The Landscape of GPU-Centric Communication. ACM Computing Surveys. 2026;58(12):1-36. doi:10.1145/3813799

BibTeX
@article{Unat2026,
  author = {Unat, Didem and Turimbetov, Ilyas and Issa, Mohammed and Sagbili, Dogan and Vella, Flavio and Sensi, Daniele and Ismayilov, Ismayil},
  title = {The Landscape of GPU-Centric Communication},
  journal = {ACM Computing Surveys},
  year = {2026},
  volume = {58},
  number = {12},
  pages = {1--36},
  doi = {10.1145/3813799}
}
Kostenlos starten →