America develops computer simulating human brain: Learns like a real person, consumes less energy than AI

American scientists have just created a prototype neuromorphic computer that can learn and remember like the human brain, paving the way for an era of energy-saving AI.

Dr. Joseph S. Friedman, Associate Professor of Electrical and Computer Engineering, The University of Texas at Dallas (UT Dallas) - Photo: UT Dallas

According to EurekAlert! on October 30, scientists at the University of Texas at Dallas (UT Dallas, USA) have developed a prototype "brain-simulating computer" capable of learning and predicting patterns with less training and energy than conventional AI systems.

This is a major step forward in the field of neurocomputing – technology inspired by the way the human brain processes and stores information.

The work, led by Dr. Joseph S. Friedman, was published in the journal Nature Communications Engineering, in collaboration with Everspin Technologies and Texas Instruments.

Unlike traditional computers that separate memory and processing, neuromorphic computers combine these two functions in the same system, making them more efficient and energy-saving.

The device operates on the principle that "neurons that work together will connect more strongly", simulating the mechanism of memory formation and learning in the human brain.

The team's main focus is on using "magnetic tunnel junctions" (MTJs) - tiny electrically adjustable components like synapses - that allow the machine to "learn" by changing the connections between artificial neurons, similar to how the human brain adapts when learning.

The project is considered a promising direction to replace current energy-consuming AI models. The research received funding from the US National Science Foundation (NSF) and the US Department of Energy, with a total budget of nearly 500,000 USD over two years to expand the experiment.

Back to topic

PUBLIC REVELATION

Source: https://tuoitre.vn/my-phat-trien-may-tinh-mo-phong-nao-nguoi-hoc-nhu-nguoi-that-it-ton-nang-luong-hon-ai-20251103085615027.htm