EUVIMEDEUVIMED entdecken
Crossref

hls4ml: A Flexible, Open Source Platform for Deep Learning Acceleration on Reconfigurable Hardware

Jan-Frederik Schulte, Benjamin Ramhorst, Chang Sun, Jovan Mitrevski, Nicolò Ghielmetti, Enrico Lupi, Dimitrios Danopoulos, Vladimir Lončar, Javier Duarte, David Burnette, Lauri Laatu, Stylianos Tzelepis und 28 weitere

ACM Transactions on Reconfigurable Technology and Systems · 30.06.2026 · 19 (2) · 1-35

DOI: 10.1145/3801979 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Transactions on Reconfigurable Technology and Systems, online erschienen am 27.05.2026. Verfasst von Jan-Frederik Schulte und 39 weiteren Autor:innen. Die Arbeit wurde bisher 13-mal zitiert (Crossref) und nennt 136 Literaturangaben. Gefördert von: Accelerated AI Algorithms for Data-Driven Discovery (A3D3) Institute under U.S. National Science Foundation, FermiForward Discovery Group, LLC, Science and Technology Facilities Council.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We present hls4ml , a free and open source platform that translates machine learning (ML) models from modern deep learning frameworks into high-level synthesis (HLS) code that can be integrated into full designs for field-programmable gate arrays (FPGAs) or application-specific integrated circuits (ASICs). With its flexible and modular design, hls4ml supports a large number of deep learning frameworks and can target HLS compilers from several vendors, including Vitis HLS, Intel oneAPI and Catapult HLS. Together with a wider eco-system for software-hardware co-design, hls4ml has enabled the acceleration of ML inference in a wide range of commercial and scientific applications where low latency, resource usage, and power consumption are critical. In this article, we describe the structure and functionality of the hls4ml platform. The overarching design considerations for the generated HLS code are discussed, together with selected performance results.

Förderung

  • Accelerated AI Algorithms for Data-Driven Discovery (A3D3) Institute under U.S. National Science Foundation
  • FermiForward Discovery Group, LLC
  • Science and Technology Facilities Council
  • EPSRC
  • CERN Next Generation Triggers project
  • European Research Council
  • DOE ASRSP GCFA
  • NSF POSE Phase II Award

Zitieren

APA

Schulte, J., Ramhorst, B., Sun, C., Mitrevski, J., Ghielmetti, N., Lupi, E., Danopoulos, D., Lončar, V., Duarte, J., Burnette, D., Laatu, L., Tzelepis, S., Axiotis, K., Berthet, Q., Wang, H., White, P., Demirsoy, S., Colombo, M., Aarrestad, T. K., …, & Gerlach, L. (2026). hls4ml: A Flexible, Open Source Platform for Deep Learning Acceleration on Reconfigurable Hardware. ACM Transactions on Reconfigurable Technology and Systems, 19(2), 1-35. https://doi.org/10.1145/3801979

Vancouver

Schulte J, Ramhorst B, Sun C, Mitrevski J, Ghielmetti N, Lupi E, et al. hls4ml: A Flexible, Open Source Platform for Deep Learning Acceleration on Reconfigurable Hardware. ACM Transactions on Reconfigurable Technology and Systems. 2026;19(2):1-35. doi:10.1145/3801979

BibTeX
@article{Schulte2026,
  author = {Schulte, Jan-Frederik and Ramhorst, Benjamin and Sun, Chang and Mitrevski, Jovan and Ghielmetti, Nicolò and Lupi, Enrico and Danopoulos, Dimitrios and Lončar, Vladimir and Duarte, Javier and Burnette, David and Laatu, Lauri and Tzelepis, Stylianos and Axiotis, Konstantinos and Berthet, Quentin and Wang, Haoyan and White, Paul and Demirsoy, Suleyman and Colombo, Marco and Aarrestad, Thea Klaeboe and Summers, Sioni and Pierini, Maurizio and Guglielmo, Giuseppe Di and Ngadiuba, Jennifer and Campos, Javier and Hawks, Benjamin and Gandrakota, Abhijith and Fahim, Farah and Tran, Nhan and Constantinides, George A and Que, Zhiqiang and Luk, Wayne and Tapper, Alexander and Hoang, Duc and Paladino, Noah and Harris, Philip C. and Lai, Bo-Cheng and Valentin, Manuel and Forelli, Ryan and Ogrenci, Seda and Gerlach, Lino},
  title = {hls4ml: A Flexible, Open Source Platform for Deep Learning Acceleration on Reconfigurable Hardware},
  journal = {ACM Transactions on Reconfigurable Technology and Systems},
  year = {2026},
  volume = {19},
  number = {2},
  pages = {1--35},
  doi = {10.1145/3801979}
}
Kostenlos starten →