Title :
Achieving supercomputer performane for neural net simulation with an array of digital signal processors
Author :
Muller, Urs A. ; Baumle, Bernhard ; Kohler, Peter ; Gunzinger, Anton ; Guggenbuhl, Walter
Author_Institution :
Electron. Lab., Swiss Federal Inst. of Technol., Zurich, Switzerland
Abstract :
Music, a digital signal processor (DSP)-based system with a parallel distributed-memory architecture that provides enormous computing power yet retains the flexibility of a general-purpose computer, is discussed. It is shown that Music reaches a peak performance of 2.7 Gflops at a significantly lower cost, power consumption, and space requirement than conventional supercomputers. The Music system hardware, programming, and backpropagation implementation are described.<>
Keywords :
digital signal processing chips; multiprocessor interconnection networks; neural nets; parallel architectures; parallel programming; virtual machines; 2.7 GFLOPS; Music; backpropagation; digital signal processors array; neural net simulation; parallel distributed-memory architecture; programming; supercomputer performane; Computational modeling; Computer architecture; Concurrent computing; Costs; Digital signal processors; Distributed computing; Energy consumption; Multiple signal classification; Neural networks; Supercomputers;
Journal_Title :
Micro, IEEE
