F. Belletti

Janus: An FPGA-Based System for High-Performance Scientific Computing

Janus is a modular, massively parallel, and reconfigurable FPGA-based computing system. Each Janus module has one computational core and one host. Janus is tailored to, but not limited to, the needs of a class of hard scientific applications characterized by regular code structure, unconventional data-manipulation requirements, and a few Megabits database. The authors discuss this configurable systems architecture and focus on its use for Monte Carlo simulations of statistical mechanics, as Janus performs impressively on this class of application.

Computing for LQCD: apeNEXT

apeNEXT is the latest in the APE collaborations series of parallel computers for computationally intensive calculations such as quantum chromo dynamics on the lattice. The authors describe the computer architectural choices that have been shaped by almost two decades of collaboration activity.

Nonequilibrium spin-glass dynamics from picoseconds to a tenth of a second.

We study numerically the nonequilibrium dynamics of the Ising spin glass, for a time spanning 11 orders of magnitude, thus approaching the experimentally relevant scale (i.e., seconds). We introduce novel analysis techniques to compute the coherence length in a model-independent way. We present strong evidence for a replicon correlator and for overlap equivalence. The emerging picture is compatible with noncoarsening behavior.

An In-Depth View of the Microscopic Dynamics of Ising Spin Glasses at Fixed Temperature

Using the special-purpose computer Janus, we follow the nonequilibrium dynamics of the Ising spin glass in three dimensions for eleven orders of magnitude. The use of integral estimators for the coherence and correlation lengths allows us to study dynamic heterogeneities and the presence of a replicon mode and to obtain safe bounds on the Edwards-Anderson order parameter below the critical temperature. We obtain good agreement with experimental determinations of the temperature-dependent decay exponents for the thermoremanent magnetization. This magnitude is observed to scale with the much harder to measure coherence length, a potentially useful result for experimentalists. The exponents for energy relaxation display a linear dependence on temperature and reasonable extrapolations to the critical point. We conclude examining the time growth of the coherence length, with a comparison of critical and activated dynamics.

Ianus: an adaptive FPGA computer

With Ianus, a next-generation field-programmable gate array (FPGA)-based machine, the authors hope to build a system that can fully exploit the performance potential of FPGA devices. A software platform that simplifies Ianus programming will extend its intended application range to a wide class of interesting and computationally demanding problems.

QCD on the Cell Broadband Engine

We evaluate IBMs Enhanced Cell Broadband Engine (BE) as a possible building block of a new generation of lattice QCD machines. The Enhanced Cell BE will provide full support of double-precision floating-point arithmetics, including IEEE-compliant rounding. We have developed a performance model and applied it to relevant lattice QCD kernels. The performance estimates are supported by micro- and application-benchmarks that have been obtained on currently available Cell BE-based computers, such as IBM QS20 blades and PlayStation 3. The results are encouraging and show that this processor is an interesting option for lattice QCD applications. For a massively parallel machine on the basis of the Cell BE, an application-optimized network needs to be developed.

Monte Carlo simulations of spin glass systems on the cell broadband engine

We implement a Monte Carlo algorithm for spin glass systems and optimize for the Cell-BE processor, assessing the effectiveness of this architecture for state-of-the-art simulations. Recent developments in many-core processor architectures, like the IBM Cell BE seem to open new opportunities in this field, where computational requirements are so demanding that state-of-the-art simulations often use dedicated computing systems. We present our implementation, analyze results and compare performances with those of both commodity processors and dedicated systems.

Nonequilibrium spin glass dynamics with Janus

The out of equilibrium evolution for an Edwards‐Anderson spin glass is followed for a tenth of a second, a long enough time to let us make safe predictions about the behaviour at experimental scales. This work has been made possible by Janus, an FPGA based special purpose computer. We have thoroughly studied the spin glass correlation functions and the growth of the coherence length for L = 80 lattices in 3D. Our main conclusion is that these spin glasses follow noncoarsening dynamics, at least up to the experimentally relevant time scales.

The apeNEXT project

Numerical simulations in theoretical high-energy physics (Lattice QCD) require huge computing resources. Several generations of massively parallel computers optimised for these applications have been developed within the APE (array processor experiment) project. Large prototype systems of the latest generation, apeNEXT, are currently being assembled and tested. This contribution explains how the apeNEXT architecture is optimised for Lattice QCD, provides an overview of the hardware and software of apeNEXT, and describes its new features, like the SPMD programming model and the C compiler.