Field‐programmable gate arrays and quantum Monte Carlo: Power efficient coprocessing for scalable high‐performance computing. Issue 12 (21st February 2019)

• Record Type: Journal Article
• Authors: Cardamone, Salvatore; Kimmitt, Jonathan R. R.; Burton, Hugh G. A.; Todman, Timothy J.; Li, Shurui; Luk, Wayne; Thom, Alex J. W.
• Other Names: Alexeev Yuri guestEditor.; Harrison Robert guestEditor.
• Is Part Of: International journal of quantum chemistry. Volume 119:Issue 12(2019)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2019-02-21
• Subjects: FPGA -- quantum Monte Carlo -- variational Monte Carlo; Quantum chemistry -- Periodicals; 541.28
• DOI: 10.1002/qua.25853
• Languages: English
• ISSNs: 0020-7608

Abstract

Abstract: Massively parallel architectures offer the potential to significantly accelerate an application relative to their serial counterparts. However, not all applications exhibit an adequate level of data and/or task parallelism to exploit such platforms. Furthermore, the power consumption associated with these forms of computation renders "scaling out" for exascale levels of performance incompatible with modern sustainable energy policies. In this work, we investigate the potential for field‐programmable gate arrays (FPGAs) to feature in future exascale platforms, and their capacity to improve performance per unit power measurements for the purposes of scientific computing. We have focused our efforts on variational Monte Carlo, and report on the benefits of coprocessing with a FPGA relative to a purely multicore system. Abstract : Certain applications in quantum chemistry, such as quantum Monte Carlo, are particularly amenable to modern high‐performance computing platforms. However, the power consumption associated with distributed memory systems and graphical processing units does not align with modern energy policies. The porting of such applications to power‐efficient hardware platforms, such as field‐programmable gate arrays, requires a significant effort on the part of the software developer. The work detailed here enumerates the benefits of transitioning to energy‐efficient platforms.