The New Initiative’s Goal Is Brain-Like Computing

The Department of Energy-funded Energy Frontier Research Center will be run by Texas A&M.

 

The underlying capabilities of conventional computers are being approached as society transforms to an information-focused society. The demand for faster, more energy-efficient computing is rising quickly.

 

The first Energy Frontier Research Center financed by the Department of Energy (DOE) will be headed by Texas A&M University (EFRC). The Reconfigurable Electronic Materials Inspired by Nonlinear Neuron Dynamics (REMIND) initiative, which aims to change computing so that it operates more like the human brain for quick and effective processing, will be the center’s primary area of interest.

 

R. The director and associate director of the EFRC, respectively, will be Sarbajit Banerjee, professor in the departments of materials science and engineering and chemistry, and Stanley Williams, professor in the department of electrical and computer engineering.

 

The future of computing is at a crossroads, according to Banerjee. “Networked grids, tailored medicine, and self-driving cars are all growing, and they all use a lot of energy. To meet the demands of society, a whole new strategy centred on brain-like computing is required.

 

Modern computers are excellent at many crucial tasks, including calculating equations and performing high-precision math. However, they struggle with fundamental human skills like instantaneous learning, idea recognition, and decision-making.

 

Human brains contain nerve cells (neurons) that continuously compare incoming stimuli with previously learnt material, enabling this capacity for information processing. Through connections known as synapses, which serve as memory repositories, neurons communicate with one another using electrical and chemical impulses. Even while individual biological processes are slower than those in transistors, huge numbers of them function at once to accomplish complex computing with energy efficiency orders of magnitude greater than the most sophisticated electronic computers.

 

Banerjee remarked, “Let’s pretend we are gazing at a picture of a dog. “The human brain can nearly instantly identify the dog, its breed, and its approximate age. A computer will struggle with the fundamental recognition and may make a huge mistake while also requiring considerably more energy to do so.”

 

By designing, producing, and assembling materials that exhibit tunable nonlinear responses to incoming electrical signals, such as thresholding, amplification, integration, and embedded memory, the REMIND initiative’s researchers are learning how to replicate human neurons and synapses in electrical circuits. In other words, they are trying to build a highly effective and competent computer by simulating the processing mechanism of the human brain.

 

Williams stated that “Our centre attempts to elucidate the fundamental science of artificial neurons and synapses.” “We are eager to address a generational challenge that will alter the course of computing in the future.”

 

Their results could drastically lower the amount of energy that computers use if these strategies are successfully implemented, aiding in the fight against the energy crisis.

 

The goal, according to Banerjee, is to enable computers to learn from experience and make decisions in dynamic, uncertain contexts. The underlying chemistries and materials for creating the next generation of brain-like computing are being discovered.

 

In order to create innovations that can revolutionise energy production and cut harmful emissions, the DOE has announced the EFRC awards. In 34 states, 11 national labs and 54 universities will participate in the research projects.

 

The National Renewable Energy Laboratory, the Department of Chemistry, the Texas A&M Engineering Experiment Station, the Lawrence Berkeley National Laboratory, and Sandia National Laboratories are all partners in the EFRC. Williams, Raymundo Arroyave, Matt Pharr, Xiaofeng Qian, and Patrick Shamberger from the department of materials science and engineering, Perla Balbuena from the Artie McFerrin Department of Chemical Engineering, Marcetta Darensbourg, and Kim Dunbar from the department of chemistry are also REMIND researchers in addition to Banerjee and Williams. The group embodies the range and depth of knowledge required to address this multidisciplinary challenge.