Eur. Phys. J. Special Topics 210, 33-51 (2012)
https://doi.org/10.1140/epjst/e2012-01636-9

Review

Reconfigurable computing for Monte Carlo simulations: Results and prospects of the Janus project

¹ Departamento de Física Teórica I, Universidad Complutense, 28040 Madrid, Spain
² Instituto de Biocomputación y Física de Sistemas Complejos (BIFI), Zaragoza, Spain
³ Departamento de Física Teórica, Universidad de Zaragoza, 50009 Zaragoza, Spain
⁴ Departamento de Ingeniería Eléctrica, Electrónica y Automática, Universidad de Extremadura, 10071 Cáceres, Spain
⁵ Fundación ARAID, Diputación General de Aragón, Zaragoza, Spain
⁶ Dipartimento di Fisica, La Sapienza Università di Roma, 00185 Rome, Italy
⁷ Dipartimento di Fisica Università di Ferrara and INFN – Sezione di Ferrara, Ferrara, Italy
⁸ Dipartimento di Fisica, IPCF-CNR, UOS Roma Kerberos and INFN, La Sapienza Università di Roma, 00185 Rome, Italy
⁹ Departamento de Ingeniería, Electrónica y Comunicaciones and Instituto de Investigación en Ingeniería de Aragón (I3A), Universidad de Zaragoza, 50018 Zaragoza, Spain
¹⁰ Departamento de Física, Universidad de Extremadura, 06071 Badajoz, Spain
¹¹ Dipartimento di Matematica e Informatica Università di Ferrara and INFN – Sezione di Ferrara, Ferrara, Italy

^a Present address: Physics Department, University of Regensburg, Germany .

Received: 30 April 2012
Revised: 25 June 2012
Published online: 6 September 2012

Abstract

We describe Janus, a massively parallel FPGA-based computer optimized for the simulation of spin glasses, theoretical models for the behavior of glassy materials. FPGAs (as compared to GPUs or many-core processors) provide a complementary approach to massively parallel computing. In particular, our model problem is formulated in terms of binary variables, and floating-point operations can be (almost) completely avoided. The FPGA architecture allows us to run many independent threads with almost no latencies in memory access, thus updating up to 1024 spins per cycle. We describe Janus in detail and we summarize the physics results obtained in four years of operation of this machine; we discuss two types of physics applications: long simulations on very large systems (which try to mimic and provide understanding about the experimental non-equilibrium dynamics), and low-temperature equilibrium simulations using an artificial parallel tempering dynamics. The time scale of our non-equilibrium simulations spans eleven orders of magnitude (from picoseconds to a tenth of a second). On the other hand, our equilibrium simulations are unprecedented both because of the low temperatures reached and for the large systems that we have brought to equilibrium. A finite-time scaling ansatz emerges from the detailed comparison of the two sets of simulations. Janus has made it possible to perform spin-glass simulations that would take several decades on more conventional architectures. The paper ends with an assessment of the potential of possible future versions of the Janus architecture, based on state-of-the-art technology.