New Reservoir Computer Marks First-Ever Microelectromechanical Neural Network Application
October 18, 2018 | AIP.orgEstimated reading time: 2 minutes
As artificial intelligence has become increasingly sophisticated, it has inspired renewed efforts to develop computers whose physical architecture mimics the human brain. One approach, called reservoir computing, allows hardware devices to achieve the higher-dimension calculations required by emerging artificial intelligence. One new device highlights the potential of extremely small mechanical systems to achieve these calculations.
Image Caption: A single silicon beam (red), along with its drive (yellow) and readout (green and blue) electrodes, implements a MEMS capable of nontrivial computations. CREDIT: Guillaume Dion
A group of researchers at the Université de Sherbrooke in Québec, Canada, reports the construction of the first reservoir computing device built with a microelectromechanical system (MEMS). Published in the Journal of Applied Physics, from AIP Publishing, the neural network exploits the nonlinear dynamics of a microscale silicon beam to perform its calculations. The group’s work looks to create devices that can act simultaneously as a sensor and a computer using a fraction of the energy a normal computer would use.
The article appears in a special topic section of the journal devoted to "New Physics and Materials for Neuromorphic Computation," which highlights new developments in physical and materials science research that hold promise for developing the very large-scale, integrated "neuromorphic" systems of tomorrow that will carry computation beyond the limitations of current semiconductors today.
“These kinds of calculations are normally only done in software, and computers can be inefficient,” said Guillaume Dion, an author on the paper. “Many of the sensors today are built with MEMS, so devices like ours would be ideal technology to blur the boundary between sensors and computers.”
The device relies on the nonlinear dynamics of how the silicon beam, at widths 20 times thinner than a human hair, oscillates in space. The results from this oscillation are used to construct a virtual neural network that projects the input signal into the higher dimensional space required for neural network computing.
In demonstrations, the system was able to switch between different common benchmark tasks for neural networks with relative ease, Dion said, including classifying spoken sounds and processing binary patterns with accuracies of 78.2% and 99.9% respectively.
“This tiny beam of silicon can do very different tasks,” said Julien Sylvestre, another author on the paper. “It’s surprisingly easy to adjust it to make it perform well at recognizing words.”
Sylvestre said he and his colleagues are looking to explore increasingly complicated computations using the silicon beam device, with the hopes of developing small and energy-efficient sensors and robot controllers.
Suggested Items
ZESTRON Welcomes Whitlock Associates as New Addition to their Existing Rep Team in Florida
04/19/2024 | ZESTRONZESTRON, the leading global provider of high-precision cleaning products, services, and training solutions in the electronics manufacturing and semiconductor industries, is thrilled to announce the addition of Whitlock Associates to its esteemed network of sales representatives.
Ark Electronics Expands Global Manufacturing Factory Network in North America and Europe
04/17/2024 | PRNewswireElectronic Manufacturing Company Ark Electronics recently announced the expansion of its Global Factory Network with the addition of Electronics Manufacturing Service (EMS) capabilities in Mexico and Europe.
Phil Kinner Pushes His Limits in Engineering and Running
04/17/2024 | Linda Stepanich, IPCEngineers break things on purpose. They like the challenge of putting things back together and making them stronger and more resilient. The skills engineers use to build products can also be used to rebuild their lives after a personal loss, something Phil Kinner learned when he laced up his running shoes after many years of a sedentary lifestyle.
Keysight Accelerates Its 6G Efforts with NVIDIA 6G Research Cloud Platform
03/22/2024 | Keysight Technologies, Inc.Keysight Technologies, Inc. is collaborating on the new NVIDIA 6G Research Cloud Platform, which includes the NVIDIA Aerial Omniverse Digital Twin, an open, flexible, and interconnected framework of network emulation resources offering researchers a comprehensive tool suite on which to develop new adaptations of artificial intelligence (AI) for radio access networks (RAN).
CACI Secures $239 Million Task Order for U.S. Army Unified Network Modernization
03/22/2024 | CACI International Inc.CACI International Inc announced that it has secured a single-award technology task order worth up to $239 million with a one-year base period and four one-year option periods to modernize the U.S. Army’s Global Secure Internet Protocol Router (SIPR) Network (GSN), including the application of commercial solutions for classified (CSfC) technology to increase options for secure user access and mobility.