SpiNNaker: A multi-core System-on-Chip for massively-parallel neural net simulation
SpiNNaker: A multi-core System-on-Chip for massively-parallel neural net simulation
复制标题
DOI：
10.1109/cicc.2012.6330636
复制
发表时间：
2012-10
期刊：
影响因子：
--
通讯作者：
E. Painkras;L. Plana;J. Garside;S. Temple;Simon Davidson;J. Pepper;David M. Clark;Cameron Patterson;S. Furber
中科院分区：
文献类型：
--
作者：
E. Painkras;L. Plana;J. Garside;S. Temple;Simon Davidson;J. Pepper;David M. Clark;Cameron Patterson;S. Furber
在处理能力和通信方面，对神经元放电的大型系统的建模在计算上要求非常高。Spinnaker是一个大规模并行计算机系统，旨在实时模拟多达10亿个尖峰神经元。这台机器的基本模块是Spinnaker多核系统片上系统，这是一个全球异步本地同步(GALS)系统，有18个ARM968处理器节点驻留在同步孤岛上，周围环绕着一个轻量级的分组交换异步通信基础设施。MPSoC在一个102 mm2的芯片中包含1亿个晶体管，提供3.96GIPS的峰值性能，当所有处理器内核以额定频率运行时，在1.2V下的功耗为1W。Spinnaker芯片于2011年5月交付，完全运行，并满足功率和性能要求。
The modelling of large systems of spiking neurons is computationally very demanding in terms of processing power and communication. SpiNNaker is a massively-parallel computer system designed to model up to a billion spiking neurons in real time. The basic block of the machine is the SpiNNaker multicore System-on-Chip, a Globally Asynchronous Locally Synchronous (GALS) system with 18 ARM968 processor nodes residing in synchronous islands, surrounded by a light-weight, packet-switched asynchronous communications infrastructure. The MPSoC contains 100 million transistors in a 102 mm2 die, provides a peak performance of 3.96 GIPS and has a power consumption of 1W at 1.2V when all processor cores operate at nominal frequency. SpiNNaker chips were delivered in May 2011, were fully operational, and met power and performance requirements.