SpiNNaker: Event-Based Simulation—Quantitative Behavior:Event-based simulation - quantitative behaviour
| Autor: | Brown, Andrew D., Chad, John E., Kamarudin, Raihaan, Dugan, Kier J., Furber, Stephen B. |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2017 |
| Předmět: |
Brain modeling
Neurons real-time simulation Computational modeling neural system simulation Event-based computing neuromorphic computing specialised simulation platforms Hardware Control and Systems Engineering Hardware and Architecture Computer architecture Engines Real-time systems Information Systems |
| Zdroj: | Brown, A D, Chad, J E, Kamarudin, R, Dugan, K J & Furber, S B 2017, ' SpiNNaker: Event-Based Simulation—Quantitative Behavior : Event-based simulation-quantitative behaviour ', IEEE Transactions on Multi-Scale Computing Systems, vol. 4, no. 3, pp. 450-462 . https://doi.org/10.1109/TMSCS.2017.2748122 |
| DOI: | 10.1109/TMSCS.2017.2748122 |
| Popis: | SpiNNaker (Spiking Neural Network Architecture) is a specialised computing engine, intended for real-time simulation of neural systems. It consists of a mesh of 240x240 nodes, each containing 18 ARM9 processors: over a millioncores, communicating via a bespoke network. Ultimately, the machine will support the simulation of up to a billion neurons in real time, allowing simulation experiments to be taken to hitherto unattainable scales. The architecture achieves this by ignoring three of the axioms of computer design: the communication fabric is non-deterministic; there is no global core synchronisation, and the system state - held in distributed memory - is not coherent. Time models itself: there is no notion of computed simulation time - wallclock time is simulation time. Whilst these design decisions are orthogonal to conventional wisdom, they bring the engine behaviour closer to its intended simulation target - neural systems. We describe how SpiNNaker simulateslarge neural ensembles; we provide performance figures and outline some failure mechanisms. SpiNNaker simulation time scales 1:1 with wallclock time at least up to 9 million synaptic connections on a 768 core subsystem (~1400th of the full system) to accurately produce logically predicted results. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|