M. Baity-Jesi
ABSTRACT
We describe the past and future of the Janus project. The collaboration started in 2006 and deployed in early 2008 the Janus supercomputer, a facility that allowed to speed-up Monte Carlo Simulations of a class of model glassy systems and provided unprecedented results for some paradigms in Statistical Mechanics. The Janus Supercomputer was based on state-of-the-art FPGA technology, and provided almost two order of magnitude improvement in terms of cost/performance and power/performance ratios. More than four years later, commercial facilities are closing-up in terms of performance, but FPGA technology has largely improved. A new generation supercomputer, Janus2, will be able to improve by more than one orders of magnitude with respect to the previous one, and will accordingly be again the best choice in Monte Carlo simulations of Spin Glasses for several years to come with respect to commercial solutions.
- M. Mézard, G. Parisi and M.A. Virasoro, Spin Glass Theory and Beyond (World Scientific, Singapore, 1987); A. P. Young (editor), Spin Glasses and Random Fields, (World Scientific, Singapore, 1998).Google Scholar
- The Janus Collaboration, Phys. Rev. Lett. 101, (2008) 157201Google Scholar
- The Janus Collaboration, Comp. Phys. Comm. 178, (2008) 208-216; IANUS: Scientific Computing on an FPGA-based Architecture, in Proceedings of ParCo2007, Parallel Computing: Architectures, Algorithms and Applications (NIC Series Vol. 38, 2007) 553-560; Computing in Science & Engineering 8, (2006) 41-49; Computing in Science & Engineering 11, (2009) 48-58.Google ScholarCross Ref
- S. F. Edwards and P. W. Anderson, J. Phys. F: Metal Phys. 5, (1975) 965-974; ibid. 6, (1976) 1927-1937.Google ScholarCross Ref
- A.D. Sokal, Functional Integration: Basics and Applications (1996 Cargese School) ed. C. DeWitt-Morette, P. Cartier and A. Folacci (1997 New York: Plenum)Google Scholar
- G. Parisi and F. Rapuano, Phys. Lett. B 157, (1985) 301-302.Google ScholarCross Ref
- The Janus Collaboration: J. Stat. Phys. 135, 1121-1158 (2009); Phys. Rev. B 79, 184408 (2009); J. Stat. Mech. (2010) P05002; J. Stat. Mech. (2010) P06026; Phys. Rev. Lett. 105, 177202 (2010); Phys. Rev. B 84, 174209 (2011); Proc. Natl. Acad. Sci. USA (2012) 109, 6452-6456Google ScholarCross Ref
- Hukushima, K., Nemoto, K. Exchange Monte Carlo Method and Application to Spin Glass Simulations J Physical Soc Japan 65,1604--1608 (1995), arXiv:cond-mat/9512035, E. Marinari, Optimized Monte Carlo Methods in Advances in Computer Simulation. Lecture Notes in Physics, 1998, Volume 501/1998, 50-81 (1996), arXiv:cond-mat/9612010.Google Scholar
- M. Guidetti et al., Spin Glass Monte Carlo Simulations on the Cell Broadband Engine in Proc. of PPAM09, (Lecture Notes on Computer Science (LNCS) 6067, Springer 2010) 467-476. M. Guidetti et al., Monte Carlo Simulations of Spin Systems on Multi-core Processors (Lecture Notes on Computer Science (LNCS) 7133 K. Jonasson (ed.), Springer, Heidelberg 2010) 220-230. Google ScholarDigital Library
Index Terms
- Janus2: an FPGA-based supercomputer for spin glass simulations
Recommendations
Scalable mpNoC for massively parallel systems - Design and implementation on FPGA
The high chip-level integration enables the implementation of large-scale parallel processing architectures with 64 and more processing nodes on a single chip or on an FPGA device. These parallel systems require a cost-effective yet high-performance ...
An FPGA implementation for neural networks with the FDFM processor core approach
This paper presents a field programmable gate array FPGA implementation of a three-layer perceptron using the few DSP blocks and few block RAMs FDFM approach implemented in the Xilinx Virtex-6 family FPGA. In the FDFM approach, multiple processor cores ...
Comments