Engineers develop a revolutionary “Brain Like Computer” at UT Dallas that learns and adapts like the human brain, redefining AI efficiency and power.
“Brain Like Computer” Sets New Benchmark in AI Innovation
In a major scientific leap, engineers at the University of Texas at Dallas have created a “Brain Like Computer” that learns, adapts, and processes information much like the human brain. This neuromorphic prototype marks a transformative step in energy-efficient artificial intelligence and could reshape how computers think, learn, and conserve power.
The research, published in Nature Communications Engineering, introduces a system built using magnetic tunnel junctions (MTJs)—microscopic devices that mimic biological synapses. By allowing the system to strengthen or weaken connections dynamically, the Brain Like Computer demonstrates Hebbian learning, the same principle that governs human neural adaptation.
Must Read: Nvidia-Samsung AI Partnership Launches Global Megafactory to Transform Semiconductor Manufacturing
How the “Brain Like Computer” Works
Unlike traditional computers that separate memory and processing units, the Brain Like Computer integrates both functions, emulating the brain’s architecture. This integration minimizes data transfer needs and drastically boosts energy efficiency.
Led by Dr. Joseph S. Friedman in collaboration with Everspin Technologies and Texas Instruments, the research team achieved over 600 trillion operations per second per watt—far surpassing the efficiency of existing AI hardware.
This advancement bridges the gap between biological intelligence and digital computation, offering machines that can “learn by doing” instead of relying on massive datasets or energy-intensive training.
Dr. Joseph S. Friedman and his colleagues created a computer prototype that learns patterns and makes predictions using fewer training computations than conventional artificial intelligence systems.
Industry Collaboration and Federal Funding Boost
The Brain Like Computer project has attracted significant industrial and governmental attention. With Everspin Technologies and Texas Instruments joining forces with UT Dallas, the research aims to accelerate commercial deployment in devices like smartphones, wearables, and smart sensors.
Dr. Sanjeev Aggarwal, CEO of Everspin Technologies and co-author of the study, emphasized that this collaboration could bring neuromorphic computing to everyday electronics.
In September 2025, the U.S. Department of Energy granted Dr. Friedman $498,730 over two years to further develop this Brain Like Computer technology. This funding highlights the growing need to curb the soaring energy demands of AI systems and create sustainable computation solutions.
Must Read: Scientists Dispel the Long-Standing Exercise Myth About Heartbeats and Longevity
Performance and Real-World Impact
In experimental trials, the Brain Like Computer achieved remarkable results—scoring 90% accuracy in handwritten digit recognition while consuming only a fraction of the energy used by traditional GPUs.
Conventional AI models often consume electricity equivalent to hundreds of homes annually. In contrast, the human brain operates on the energy of a light bulb. The new Brain Like Computer promises up to 80% lower power consumption, addressing both ecological and economic concerns.
This could enable advanced AI applications on portable or remote devices without relying on cloud processing—ensuring faster, private, and energy-efficient decision-making.
Societal Implications and the Future of AI
As this Brain Like Computer moves from labs to real-world systems, researchers are focused on scalability, reliability, and ethical AI deployment. The technology could democratize access to artificial intelligence by making it affordable and sustainable, but it also raises questions about privacy, data control, and human-AI coexistence.
Experts believe that neuromorphic computing could redefine industries—from autonomous vehicles to healthcare—by enabling machines that not only analyze but understand data patterns in real time.
The UT Dallas team’s work represents more than a technical triumph; it’s a philosophical leap toward creating machines that mimic cognitive learning—machines that think like us.
The Road Ahead for Brain Like Computer Technology
The next phase of the Brain Like Computer project involves scaling up chip architecture, integrating it into complex systems, and enhancing its robustness for long-term use. Future applications could include AI assistants that adapt intuitively to user behavior, medical devices that analyze neural activity, and edge systems that process vast data with minimal energy use.
With industry partnerships, federal backing, and a clear path toward commercialization, this Brain Like Computer innovation stands at the frontier of a new computing era—where intelligence, adaptability, and efficiency converge seamlessly.