Ultra low power of artificial cognitive memory for brain-like computation

The development of conventional computer is limited by the separated CPU and memory, as well as the desperate Moore's law. It's a trend to develop a brain-like computer architecture in order to solve above problems, where a new memory is highly needed. Artificial cognitive memory (ACM) based on neuromorphic circuits, integrates computation tightly with storage like the brain and is the key component to support brain-like computation. The power issue of ACM is critical, especially for the application of portable devices. In this paper, we systematically analyze the key factors of the energy consumption for building an ACM chip, such as element model, silicon technology, synaptic material, array selecting, weight modification method, routing strategy and the integration approaches. These results provide an in-depth insight for creating an ultra-low power brain-like computing platform.

[1]  Byoungil Lee,et al.  Nanoelectronic programmable synapses based on phase change materials for brain-inspired computing. , 2012, Nano letters.

[2]  Shimeng Yu,et al.  An Electronic Synapse Device Based on Metal Oxide Resistive Switching Memory for Neuromorphic Computation , 2011, IEEE Transactions on Electron Devices.

[3]  Geoffrey W. Burr,et al.  Nanoscale electronic synapses using phase change devices , 2013, JETC.

[4]  Rodrigo Alvarez-Icaza,et al.  Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations , 2014, Proceedings of the IEEE.

[5]  Eugene M. Izhikevich,et al.  Which model to use for cortical spiking neurons? , 2004, IEEE Transactions on Neural Networks.

[6]  Paul E. Hasler,et al.  Floating Gate Synapses With Spike-Time-Dependent Plasticity , 2011, IEEE Transactions on Biomedical Circuits and Systems.

[7]  Eugenio Culurciello,et al.  Capacitive Inter-Chip Data and Power Transfer for 3-D VLSI , 2006, IEEE Transactions on Circuits and Systems II: Express Briefs.

[8]  Scott Koziol,et al.  Low Power Dendritic Computation for Wordspotting , 2013 .

[9]  Luping Shi,et al.  Artificial cognitive memory—changing from density driven to functionality driven , 2011 .

[10]  Narayan Srinivasa,et al.  A functional hybrid memristor crossbar-array/CMOS system for data storage and neuromorphic applications. , 2012, Nano letters.

[11]  Jim D. Garside,et al.  SpiNNaker: A 1-W 18-Core System-on-Chip for Massively-Parallel Neural Network Simulation , 2013, IEEE Journal of Solid-State Circuits.

[12]  Wei Yang Lu,et al.  Nanoscale memristor device as synapse in neuromorphic systems. , 2010, Nano letters.

[13]  Andrew S. Cassidy,et al.  A million spiking-neuron integrated circuit with a scalable communication network and interface , 2014, Science.

[14]  Eric Pop,et al.  Low-Power Switching of Phase-Change Materials with Carbon Nanotube Electrodes , 2011, Science.

[15]  Jennifer Hasler,et al.  Finding a roadmap to achieve large neuromorphic hardware systems , 2013, Front. Neurosci..

[16]  L. Abbott,et al.  Competitive Hebbian learning through spike-timing-dependent synaptic plasticity , 2000, Nature Neuroscience.