We study the properties of spatially localized and time-periodic excitations--discrete breathers--in Fermi-Pasta-Ulam (FPU) chains. We provide a detailed analysis of their spatial profiles and stability properties. We especially demonstrate that the Page mode is linearly stable for symmetric FPU potentials. A resonant interaction between a localized and delocalized perturbations causes weak but...

We address the question of the effect of disorder on heat conduction in an anharmonic chain with interactions given by the Fermi-Pasta-Ulam (FPU) potential. In contrast to the conclusions of an earlier paper [Phys. Rev. Lett. 86, 63 (2001)], which found that disorder could induce a finite thermal conductivity at low temperatures, we find no evidence of a finite-temperature transition in conduct...

Some exact solutions and multi-mode invariant submanifolds were found for the Fermi-Pasta-Ulam (FPU) β-model by Poggi and Ruffo in Phys. D 103 (1997) 251. In the present paper we demonstrate how results of such a type can be obtained for an arbitrary Nparticle chain with periodic boundary conditions with the aid of our group-theoretical approach [Phys. D 117 (1998) 43] based on the concept of b...

The PowerPCt 440 floating-point unit (FPU) with complexarithmetic extensions is an embedded application-specific integrated circuit (ASIC) core designed to be used with the IBM PowerPC 440 processor core on the Blue Genet/L compute chip. The FPU core implements the floating-point instruction set from the PowerPC Architecturee and the floating-point instruction extensions created to aid in matri...

In the last ten years, interval techniques [1, 2] have allowed original solutions for many problems in engineering to be proposed, see, e.g., [3]. One of the main features of interval techniques is their ability to provide guaranteed results, i.e., with a verified accuracy or which are numerically proved. Consider for example, a bounded-error parameter estimation problem: the value of some para...

This paper is devoted to a numerical study of the familiar α+β FPU model. Precisely, we here discuss, revisit and combine together two main ideas on the subject: (i) In the system, at small specific energy ε = E/N , two well separated time–scales are present: in the former one a kind of metastable state is produced, while in the second much larger one, such an intermediate state evolves and rea...

Floating-point arithmetic units (FPU) have paramount importance in applications that involve intensive mathematic operations. However, previous implementations of FPU either require much manual work or only support special functions (e.g. reciprocal, square root, logarithm, etc.). In this paper, we present an automatic method to synthesize general FPU by aligned partition. Based on the novel pa...

In this work, we provide energy-efficient architectural support for floating point accuracy. Our goal is to provide accuracy that is far greater than that provided by the processor’s hardware floating point unit (FPU). Specifically, for each floating point addition performed, we “recycle” that operation’s error: the difference between the finite-precision result produced by the hardware and the...