CoChat

The Future of FPGA Acceleration in Datacenters and the Cloud

The DOI matches its Crossref record; 139 of the 143 references it cites were found in a registry or an index.

Abstract

In this article, we survey existing academic and commercial efforts to provide Field-Programmable Gate Array (FPGA) acceleration in datacenters and the cloud. The goal is a critical review of existing systems and a discussion of their evolution from single workstations with PCI-attached FPGAs in the early days of reconfigurable computing to the integration of FPGA farms in large-scale computing infrastructures. From the lessons learned, we discuss the future of FPGAs in datacenters and the cloud and assess the challenges likely to be encountered along the way. The article explores current architectures and discusses scalability and abstractions supported by operating systems, middleware, and virtualization. Hardware and software security becomes critical when infrastructure is shared among tenants with disparate backgrounds. We review the vulnerabilities of current systems and possible attack scenarios and discuss mitigation strategies, some of which impact FPGA architecture and technology. The viability of these architectures for popular applications is reviewed, with a particular focus on deep learning and scientific computing. This work draws from workshop discussions, panel sessions including the participation of experts in the reconfigurable computing field, and private discussions among these experts. These interactions have harmonized the terminology, taxonomy, and the important topics covered in this manuscript.

Overview

Integrity signals

Registry

Crossref · 10.1145/3506713

Cites

139 of 139 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Verification

Claims the paper makes

This paper’s full text isn’t openly available, so its claims and self-check weren’t run. The registry and citation checks still apply.

Citations

Cites

Research works printed but not linked: 35

  1. 10.5555/874060.875374 · doi:10.5555/874060.875374
  2. ARM. 2010. AMBA 4 AXI4-Stream Protocol Specification. Technical Report. ARM.
  3. ARM. 2011. AMBA AXI and ACE Protocol Specification. Technical Report. ARM.
  4. Michaela Blott, Kimon Karras, Ling Liu, Kees Vissers, Jeremia Bär, and Zsolt István. 2013. Achieving 10Gbps line-rate key-value stores with FPGAs. In 5th USENIX Workshop on Hot Topics in Cloud Computing (HotCloud’13). USENIX Association, San Jose, CA. Retrieved from https://www.usenix.org/conference/hotcloud13/workshop-program/presentations/blott.
  5. 10.5555/2813749.2813757 · doi:10.5555/2813749.2813757
  6. 10.5555/1564913 · doi:10.5555/1564913
  7. 10.5555/2650280.2650362 · doi:10.5555/2650280.2650362
  8. 10.5555/3195638.3195647 · doi:10.5555/3195638.3195647
  9. 10.5555/3027041.3027052 · doi:10.5555/3027041.3027052
  10. Chris Conger, Ian Troxel, D. Espinoza, Vikas Aggarwal, and A. George. 2005. NARC: Network attached reconfigurable computing for high performance, network based applications. In 8th Annual International Conference on Military and Aerospace Programmable Logic Devices (MAPLD’05).
  11. Maik Ender, Amir Moradi, and Christof Paar. 2020. The unpatchable silicon: A full break of the bitstream encryption of Xilinx 7-series FPGAs. In 29th USENIX Security Symposium.
  12. 10.5555/3358807.3358838 · doi:10.5555/3358807.3358838
  13. Dieter Ernst. 2020. Competing in Artificial Intelligence Chips: China’s Challenge Amid Technology War. Centre for International Governance Innovation, Special Report.
  14. 10.5555/3307441.3307446 · doi:10.5555/3307441.3307446
  15. William Gayde. 2020. How Arm Came to Dominate the Mobile Market and It’s Coming for More, Much More. Retrieved from https://www.techspot.com/article/1989-arm-inside.
  16. Nicole Hemsothd. 2017. First in-depth look at Google’s new second-generation TPU. Retrieved from https://www.nextplatform.com/2017/05/17/first-depth-look-googles-new-second-generation-tpu/.
  17. 10.5555/2930611.2930639 · doi:10.5555/2930611.2930639
  18. 10.5555/3488766.3488822 · doi:10.5555/3488766.3488822
  19. 10.5555/3277203.3277276 · doi:10.5555/3277203.3277276
  20. Peter Loscocco and Stephen Smalley. 2001. Meeting critical security objectives with security-enhanced Linux. In Ottawa Linux Symposium. 115–134.
  21. Layong Larry Luo and T. E. G. Tencent. 2018. In Towards Converged SmartNIC Architecture for Bare Metal and Public Clouds at Tencent Scale 2nd Asia-Pacific Workshop on Networking (APNet 2018) August 2-3 2018 Beijing China .
  22. 10.5555/2971808.2972016 · doi:10.5555/2971808.2972016
  23. Christian Plessl. 2018. Bringing FPGAs to HPC production systems and codes. In H2RC’18 Workshop at Supercomputing (SC’18). DOI:https://doi.org/10.13140/RG.2.2.34327.42407 · doi:10.13140/rg.2.2.34327.42407
  24. 10.5555/647325.721663 · doi:10.5555/647325.721663
  25. Grigory Sapunov. 2021. Will ASIC Chips Become the Next Big Thing in AI? Retrieved from https://moorinsightsstrategy.com/will-asic-chips-become-the-next-big-thing-in-ai/.
  26. 10.5555/3195638.3195659 · doi:10.5555/3195638.3195659
  27. 10.5555/1204009 · doi:10.5555/1204009
  28. 10.5555/2978159 · doi:10.5555/2978159
  29. 10.5555/3437539.3437757 · doi:10.5555/3437539.3437757
  30. Naif Tarafdar and Paul Chow. 2019. libGalapagos: A software environment for prototyping and creating heterogeneous FPGA and CPU applications. In 6th International Workshop on FPGAs for Software Programmers (FSP’19).
  31. Malte Vesper, Dirk Kocha, and Khoa Phama. 2017. PCIeHLS: An OpenCL HLS framework. In 4th International Workshop on FPGAs for Software Programmers. VDE, 1–6.
  32. Greg Watson, Nick McKeown, and Martin Casado. 2006. NetFPGA: A tool for network research and education. In 2nd Workshop on Architectural Research Using FPGA Platforms (WARFP), Vol. 3.
  33. Xilinx Case Study. [n.d.]. Xilinx Powers Alibaba Cloud FaaS with AI Acceleration Solution for E-Commerce Business. Retrieved from https://www.xilinx.com/publications/powered-by-xilinx/xilinx-alibaba-case-study.pdf.
  34. Xilinx Corporation 2021. Virtex UltraScale+ FPGA Data Sheet: DC and AC Switching Characteristics.
  35. Wang Xu. 2018. Hardware acceleration over NFV in China Mobile. OPNFV Plugfest.

Cited by

See all 158

Metadata

Authors
Bobda, C., Mbongue, J. M., Chow, P., Ewais, M., Tarafdar, N., Vega, J. C., Eguro, K., Koch, D., Handagala, S., Leeser, M., Herbordt, M., Shahzad, H., Hofste, P., Ringlein, B., Szefer, J., Sanaullah, A., & Tessier, R.
Affiliations
  • University of Florida (US)
  • University of Toronto (CA)
  • Microsoft (United States) (US)
  • University of Manchester (GB)
  • Northeastern University (US)
  • Boston University (US)
  • IBM (United States) (US)
  • IBM Research - Zurich (CH)
  • Yale University (US)
  • Red Hat (United States) (US)
  • University of Massachusetts Amherst (US)
Venue
ACM Transactions on Reconfigurable Technology and Systems
Published
2022-02-04
Volume
15
Issue
3
Pages
1-42
ISSN
  • 1936-7406 (ISSN-L)
  • 1936-7414
OpenAlex
W4210258659
Cited by
158
References
143
Citations by year
2026: 21 · 2025: 43 · 2024: 43 · 2023: 40 · 2022: 11
FWCI
26.12
Type
journal article
Language
en
Open access
closed
Locations
ACM Transactions on Reconfigurable Technology and Systems (publishedVersion)
Topics
  • Physical Sciences › Computer Science › Hardware and Architecture › Embedded Systems Design Techniques
  • Physical Sciences › Computer Science › Information Systems › Cloud Computing and Resource Management
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › VLSI and FPGA Design Techniques
Keywords
FPGA acceleration, reconfigurable computing, data centers, cloud computing, scalability, virtualization, middleware, operating system abstractions, hardware security, tenant isolation, attack scenarios, deep learning acceleration, scientific computing, FPGA architecture

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.