HETEROGENEOUS SUPERCOMPUTING SYSTEMS BASED ON OPENCL SOFTWARE PLATFORM
Abstract
The increase of the demands, that society specifies to the computing systems, does not corre-sponds to the increase in their performance. This is due to the approach of the VLSI integration to its physical limit, as well as with the use of general-purpose architectures to perform highly spe-cialized tasks. The first problem is fundamental and can not be figured out in short term perspective until mass-market quantum or optical computers appear. The second one could be solved today. High performance computing world tends to move from clusters, based on general purpose processors to heterogeneous structures, such as: GPUs and ASIC. However the most perspective approach is to use hardware-reconfigurable computers based on FPGAs. They have both the ad-vantages of ASIC, that is low power consumption and high efficiency on a specific task, and the flexibility of GPU, i.e. configuration could be changed by software. The current paper shows the advantages of using reconfigurable computers instead of traditional approaches. Also it describes a unique reconfigurable computing board based on an array of 4 Xilinx FPGA. The board-support package, which allows creating configuration using OpenCL language, has been made for this board. OpenCL is a cross platform standard for high-performance parallel computing. We are integrating the board to a reconfigurable supercomputer, and developing an intellectual profiler, that is going to suppose to determine which computational unit suits better for the current task.
References
2. Antonov A.P., Zaborovskiy V.S., Kiselev I.O. Spetsializirovannye rekonfiguriruemye vychisliteli v setetsentricheskikh superkomp'yuternykh sistemakh [Specialized reconfigurable calculators in network-centric supercomputer systems], Sistemy vysokoy dostupnosti [High availability systems], 2018, Vol. 14, No. 3, pp. 57-62.
3. Awad M. FPGA supercomputing platforms: A survey, IEEE Conference Paper: 19th Interna-tional Conference on Field Programmable Logic and Applications, 2009, pp. 564-568.
4. Antonov A.P., Mamutova O.V., Filippov A.S. Next Generation FPGA-based Platform for Network Security, 18th Conference of Open Innovations Association (FRUCT), Information Security and Protection of Information Technologies Session, St. Petersburg, IEEE, 2016, pp. 116-121.
5. Kalyaev I.A., Levin I.I., Semernikov E.A., SHmoylov V.I. Razvitie otechestvennykh mnogokristal'nykh rekonfiguriruemykh vychislitel'nykh sistem: ot vozdushnogo k zhidkostnomu okhlazhdeniyu [The development of domestic multi-chip reconfigurable computing systems: from air to liquid cooling], Trudy SPIIRAN [Proceedings of SPIIRAS], 2017, No. 1, pp. 27-39.
6. OpenCL specification 1.2, Khronos Group, 2012, 380 p.
7. Kobayashi R., Oobata, Y., Fujita, N., Yamaguchi, Y., Boku, T. OpenCL-ready High Speed FPGA Network for Reconfigurable High Performance Computing, HPC Asia 2018 Proceed-ings of the International Conference on High Performance Computing in Asia-Pacific Region, 2018, pp. 192-201.
8. Zwagerman M. High Level Synthesis, a Use Case Comparison with Hardware Description Language, Grand Valley State University Masters Theses, 2015, 36 p.
9. Intel FPGA Devices. Available at: https://www.intel.com/content/www/us/en/fpga/ devic-es.html (accessed 04 November 2018).
10. Xilinx Inc. Available at: https://www.xilinx.com/ (accessed 04 November 2018).
11. Pelcat M., Cédric Bourrasset, Luca Maggiani, François Berry. Design Productivity of a High Level Synthesis Compiler versus HDL, Proceedings of the 2016 International Conference on Embedded Computer Systems: Architectures, Modeling and Simulation, 2016, pp. 140-147.
12. Shuhao Z., Bingsheng H., Wei Z. Melia a MapReduce framework on OpenCL based FPGA, IEEE Transactions on Parallel and Distributed Systems, 2015, pp. 27-39.
13. Rumyantsev Yu.A., Zakharov P.N., Abrashitova N.A., SHmatok A.V., Ryzhikh V.O., Gudimchuk N.B., Ataullakhanov F.I. Primenenie PLIS dlya rascheta depolimerizatsii mikrotrubochki metodom brounovskoy dinamiki [The use of FPGAs to calculate the depolymerization of mi-crotubules by the method of Brownian dynamics], Trudy Instituta sistemnogo programmirovaniya RAN [Proceedings of Institute for system programming of Russian Acad-emy of Sciences], 2016, Vol. 28, Issue 3, pp. 241-266.
14. Meeus W., Van Beeck K., Goedemé, T., Meel J., & Stroobandt D. An overview of today’s high-level synthesis tools, Design Automation for Embedded Systems, 2012, Vol. 16 (3), pp. 31-51.
15. Gao S.,Chirtz J. Characterization of OpenCL on a scalable FPGA architecture, 2014 Interna-tional Conference on ReConFigurable Computing and FPGAs, 8.12.2014, 8 p.
16. Hill K., Craciun S., George A., & Lam H. Comparative analysis of OpenCL vs. HDL with image-processing kernels on Stratix-V FPGA, 2015 IEEE 26th International Conference on Application-Specific Systems, Architectures and Processors (ASAP), pp. 189-193.
17. Dordopulo A.I., Levin I.I., Kalyaev I.A., Gudkov V.A., Gulenok A.A. Programmirovanie vychislitel'nykh sistem gibridnogo tipa na osnove metoda reduktsii proizvoditel'nosti [Program-ming of hybrid computing systems based on the method of performance reduction], Parallel'nye vychislitel'nye tekhnologii: Trudy Mezhdunarodnoy nauchnoy konferentsii [Parallel computing technologies: Proceedings of the International scientific conference], 2016, pp. 131-140.
18. Quentin Gautier, Alric Althoff, Pingfan Meng, Ryan Kastner Spector. An OpenCL FPGA benchmark suite, International Conference on Field-Programmable Technology (FPT), Xi'an, China, 2016, pp. 123-134.
19. Chang D.W., Jenkins C.D., Garcia P.C. et al. ERCBench: An open-source benchmark suite for embedded and reconfigurable computing, Proceedings – 2010 International Conference on Field Programmable Logic and Applications FPL, 2010, pp. 408-413.
20. Murray K.E., Whitty S., Liu S. et al. Titan: Enabling large and complex benchmarks in aca-demic cad, 2013 23rd International Conference on Field programmable Logic and Applica-tions, Sept 2013, pp. 1-8.
21. Feng W.-c., Lin H., Scogland T. et al. Opencl and the 13 dwarfs: A work in progress, Proceed-ings of the 3rd ACM/SPEC International Conference on Performance Engineering ser. ICPE '12, 2012, pp. 291-294.
22. Xilinx sdaccel. Available at: http://www.xilinx.com/products/design-tools/software-zone/sdaccel.html (accessed 04 November 2018).
23. Nauchno-tekhnologicheskaya infrastruktura Rossiyskoy Federatsii [Scientific and technologi-cal infrastructure of the Russian Federation]. Available at: http://ckp-rf.ru/ (accessed 04 No-vember 2018).
