Advertisement

Reconfigurable Computer Based on Virtex UltraScale+ FPGAs with Immersion Cooling System

  • I. I. Levin
  • A. I. DordopuloEmail author
  • A. M. Fedorov
  • A. A. Gulenok
Conference paper
Part of the Communications in Computer and Information Science book series (CCIS, volume 753)

Abstract

The paper covers the reconfigurable computer systems based on high integration field programmable gate arrays (FPGAs) of Xilinx Virtex UltraScale+ series, where immersion liquid cooling system is used for cooling of all electronic components. The original engineering and technological solutions for thermal interface, cooling liquid, radiators and the entire reconfigurable computer system are considered. Owing to these features it is possible to achieve unprecedented layout density, up to 128 FPGAs for a 3U computational module placed into a standard \(19^{{\prime }{\prime }}\) computer rack. The paper presents results of prototyping and experimental verification of energy-efficient computational module with immersion liquid cooling system. On the base of the new computational module it is possible to achieve the performance of 1 Pflops in a standard 47U rack with total power consumption not more than 150 kW. The designed immersion liquid cooling system has power reserve for already produced and for next-generation high integration FPGA series, resistance to leaks and their consequences, and compatibility with traditional water cooling systems based on industrial chillers.

Keywords

Reconfigurable computer systems Immersion liquid cooling system FPGAs Liquid cooling Computational module High-performance computer systems Real and specific performance Energy efficiency 

References

  1. 1.
    Kalyaev, I.A., Levin, I.I., Semernikov, E.A., Shmoilov, V.I.: Reconfigurable Multipipeline Computing Structures, 330 p. Nova Science Publishers, New York (2012)Google Scholar
  2. 2.
    Tripiccione, R.: Reconfigurable computing for statistical physics: the weird case of JANUS. In: IEEE 23rd International Conference on Application-specific Systems, Architectures and Processors (ASAP) (2012)Google Scholar
  3. 3.
    Baity-Jesi, M., et al.: The janus project: boosting spin-glass simulations using FPGAs. IFAC Proc. Vols., Program. Dev. Embedded Syst. 12(1) (2013)Google Scholar
  4. 4.
    Shaw, D.E., et al.: Anton, a special-purpose machine for molecular dynamics simulation. Commun. ACM 51(7), 91–97 (2008)CrossRefGoogle Scholar
  5. 5.
    Levin, I.I., Dordopulo, A.I., Doronchenko, Y.I., Raskladkin, M.K.: Reconfigurable computer system on the base of Virtex UltraScale FPGAs with liquid cooling. In: Proceedings of the 10th Annual International Scientific Conference on Parallel Computing Technologies. CEUR Workshop Proceedings, vol. 1576, pp. 221–230 (2016)Google Scholar
  6. 6.
    Shah, J.M., Eiland, R., Siddarth, A., Agonafer, D.: Effects of mineral oil immersion cooling on IT equipment reliability and reliability enhancements to data center operations. In: Proceedings of the 15th InterSociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm 2016), pp. 316–325, 20 July 2016. Article number 7517566Google Scholar
  7. 7.
    Li, L., Zheng, W., Wang, X.: Data center power minimization with placement optimization of liquid-cooled servers and free air cooling. Sustain. Comput. Inf. Syst. 11(1), 3–15 (2016)Google Scholar
  8. 8.
    Gess, J.A., Dreher, T.A., Bhavnani, S.A., Johnson, W.B.: Effect of flow guide integration on the thermal performance of high performance liquid cooled immersion server modules. In: ASME 2015 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems (InterPACK 2015), collocated with the ASME 2015 13th International Conference on Nanochannels, Microchannels and Minichannels, vol. 1 (2015)Google Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • I. I. Levin
    • 1
  • A. I. Dordopulo
    • 1
    Email author
  • A. M. Fedorov
    • 1
  • A. A. Gulenok
    • 2
  1. 1.Scientific Research Centre of Supercomputers and Neurocomputers Co. Ltd.TaganrogRussia
  2. 2.Scientific Research Institute of Multiprocessor Computer Systems at Southern Federal UniversityRostov-on-donRussia

Personalised recommendations