Cited By
View all- Hijma PHeldens SSclocco Avan Werkhoven BBal H(2023)Optimization Techniques for GPU ProgrammingACM Computing Surveys10.1145/357063855:11(1-81)Online publication date: 16-Mar-2023
Radio-Astronomical Imaging: FPGAs vs GPUs
Pages 509 - 521
Abstract
FPGAs excel in performing simple operations on high-speed streaming data, at high (energy) efficiency. However, so far, their difficult programming model and poor floating-point support prevented a wide adoption for typical HPC applications. This is changing, due to recent FPGA technology developments: support for the high-level OpenCL programming language, hard floating-point units, and tight integration with CPU cores. Combined, these are game changers: they dramatically reduce development times and allow using FPGAs for applications that were previously deemed too complex.
In this paper, we show how we implemented and optimized a radio-astronomical imaging application on an Arria 10 FPGA. We compare architectures, programming models, optimizations, performance, energy efficiency, and programming effort to highly optimized GPU and CPU implementations. We show that we can efficiently optimize for FPGA resource usage, but also that optimizing for a high clock speed is difficult. All together, we demonstrate that OpenCL support for FPGAs is a leap forward in programmability and it enabled us to use an FPGA as a viable accelerator platform for a complex HPC application.
References
[1]
ASTRON Netherlands Institute for Radio Astronomy: Image-Domain Gridding for FPGAs (2019). https://gitlab.com/astron-idg/idg-fpga
[2]
Bal H et al. A medium-scale distributed system for computer science research: infrastructure for the long term IEEE Comput. 2016 49 5 54-63
[3]
Cong, J., et al.: Understanding performance differences of FPGAs and GPUs. In: 2018 IEEE 26th International Symposium on Field-Programmable Custom Computing Machines, pp. 93–96 (2018)
[4]
de Fine Licht, J., et al.: Transformations of high-level synthesis codes for high-performance computing. Computing Research Repository (CoRR) (2018)
[5]
Jin, Z., Finkel, H.: Power and performance tradeoff of a floating-point intensive kernel on OpenCL FPGA platform, pp. 716–720 (2018)
[6]
Minhas UI, Woods R, and Karakonstantis G Voros N, Huebner M, Keramidas G, Goehringer D, Antonopoulos C, and Diniz PC Exploring functional acceleration of OpenCL on FPGAs and GPUs through platform-independent optimizations Applied Reconfigurable Computing. Architectures, Tools, and Applications 2018 Cham Springer 551-563
[7]
Muslim FB et al. Efficient FPGA implementation of OpenCL high-performance computing applications via high-level synthesis IEEE Access 2017 5 2747-2762
[8]
Romein, J.W., Veenboer, B.: PowerSensor 2: a fast power measurement tool. In: 2018 IEEE International Symposium on Performance Analysis of Systems and Software, pp. 111–113 (2018)
[9]
Treibig, J., Hager, G., Wellein, G.: LIKWID: a lightweight performance-oriented tool suite for x86 multicore environments. In: Proceedings of the International Conference on Parallel Processing, pp. 207–216 (2010)
[10]
van der Tol, S., Veenboer, B., Offringa, A.: Image domain gridding. Astron. Astrophys. 616, A27 (2018)
[11]
Veenboer, B., Petschow, M., Romein, J.W.: Image-domain gridding on graphics processors. In: Proceedings of the International Parallel and Distributed Processing Symposium, IPDPS, pp. 545–554 (2017)
[12]
Williams S, Waterman A, and Patterson D Roofline: an insightful visual performance model for multicore architectures Commun. ACM 2009 52 65-76
[13]
Won, M.S.: Meeting the performance and power imperative of the Zettabyte era with generation 10. Technical report, Intel Programmable Solutions Group (2013)
[14]
Yang, C., et al.: OpenCL for HPC with FPGAs: case study in molecular electrostatics. In: 2017 IEEE High Performance Extreme Computing Conference (HPEC), pp. 1–8 (2017)
[15]
Zohouri, H.R.: High performance computing with FPGAs and OpenCL. Ph.D. thesis, Tokyo Institute of Technology (2018)
[16]
Zohouri, H.R., et al.: Evaluating and optimizing OpenCL kernels for high performance computing with FPGAs. In: SC16: International Conference for High Performance Computing, Networking, Storage and Analysis, pp. 409–420 (2016)
Recommendations
Exploiting Parallelism on GPUs and FPGAs with OmpSs
ANDARE '17: Proceedings of the 1st Workshop on AutotuniNg and aDaptivity AppRoaches for Energy efficient HPC SystemsThis paper presents the OmpSs approach to deal with heterogeneous programming on GPU and FPGA accelerators. The OmpSs programming model is based on the Mercurium compiler and the Nanos++ runtime. Applications are annotated with compiler directives ...
Understanding Performance Differences of FPGAs and GPUs: (Abtract Only)
FPGA '18: Proceedings of the 2018 ACM/SIGDA International Symposium on Field-Programmable Gate ArraysThe notorious power wall has significantly limited the scaling for general-purpose processors. To address this issue, various accelerators, such as GPUs and FPGAs, emerged to achieve better performance and energy-efficiency. Between these two ...
General-Purpose Computing with Soft GPUs on FPGAs
Special Section on FCCM 2016 and Regular PapersUsing field-programmable gate arrays (FPGAs) as a substrate to deploy soft graphics processing units (GPUs) would enable offering the FPGA compute power in a very flexible GPU-like tool flow. Application-specific adaptations like selective hardening of ...
Comments
Information & Contributors
Information
Published In
© Springer Nature Switzerland AG 2019.
Publisher
Springer-Verlag
Berlin, Heidelberg
Publication History
Published: 26 August 2019
Qualifiers
- Article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- View Citations1Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0
Reflects downloads up to 10 Nov 2024
Other Metrics
Citations
Cited By
View all- Hijma PHeldens SSclocco Avan Werkhoven BBal H(2023)Optimization Techniques for GPU ProgrammingACM Computing Surveys10.1145/357063855:11(1-81)Online publication date: 16-Mar-2023
View Options
View options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in