Archive : 11 May 2026 год

Researchers at the University of Missouri have made a significant discovery that could help solve one of the biggest challenges facing artificial intelligence — its rapidly growing energy consumption. By studying brain-inspired neuromorphic transistors, the team has identified key design principles that could lead to much more energy-efficient AI hardware.

Key Findings

The performance of neuromorphic transistors depends heavily on the ultrathin interface between the semiconductor and the insulating layer. This interface plays a critical role in how efficiently the device can process and store information simultaneously — mimicking the way synapses work in the human brain.

Unlike traditional computers that separate memory and processing (the von Neumann architecture), neuromorphic systems combine both functions in a single device. This approach significantly reduces energy loss caused by constantly moving data between memory and processor.

Why It Matters

AI systems are becoming extremely power-hungry. Data centers running large AI models already consume massive amounts of electricity, and demand continues to grow rapidly. The human brain, by comparison, can perform complex cognitive tasks using only about 20 watts.

The Missouri researchers’ work on organic neuromorphic transistors aims to bridge this efficiency gap, potentially enabling AI hardware that learns and processes information with far lower energy requirements.

Potential Impact

This research could have major implications for:

• Future data centers and cloud AI infrastructure
• Edge AI devices with limited power budgets
• More sustainable and scalable artificial intelligence
• Next-generation neuromorphic processors

While still in the research phase, these findings provide important design guidelines for building practical brain-like chips that could help AI scale sustainably in the coming years.

Source: Electronics For You