Penn State Researchers Develop DNA-Based Memristor Operating on Under 0.1 Volts
A new memory technology combining synthetic DNA and perovskite semiconductors cuts the energy consumption of AI computing systems by a factor of 100, researchers at the Pennsylvania State University announced. Published in the journal Advanced Functional Materials, the breakthrough addresses the soaring energy demands of data centers by utilizing biological molecules as highly efficient information storage elements, telepolis.de reported.
Integration of Synthetic DNA and Perovskite Semiconductors
The newly developed memory element consists of commercially available synthetic DNA forming short genetic segments combined with perovskite, a semiconductor material already used in solar cells and lasers. To make the biological material conductive, researchers added silver nanoparticles to customized DNA sequences of specific compositions and lengths. This doped DNA layer was applied over a thin perovskite film, allowing the molecules to arrange themselves systematically while conducting electrical current.
Neuromorphic Computing and Low-Voltage Operation
The resulting component functions as a memristor—a resistor that changes its resistance based on previous electrical current and retains that state even after power is turned off. Operating reliably at voltages below 0.1 volts, the device stores and processes information in the same location rather than separating these functions like traditional computer architectures. This processing method mimics human brain neurons, advancing neuromorphic computing systems designed for complex artificial intelligence workloads.
Performance Stability and Industrial Scaling Challenges
Laboratory tests demonstrated that the DNA-perovskite memristors remained stable for six weeks at temperatures reaching 120 degrees Celsius, outperforming standard perovskite-based storage elements. Bed Poudel, a materials scientist at Penn State and corresponding author of the study, noted that the device achieves higher storage capacity than conventional flash drives while consuming a fraction of the power. However, widespread industrial adoption depends on overcoming hurdles related to production costs, long-term durability, scalability, and compatibility with existing storage architectures.
Frequently Asked Questions About DNA-Based AI Memory
How much data can a single gram of DNA store?
A single gram of DNA can store approximately 215 million gigabytes of digital information, according to the research data.
When was the concept of DNA data storage first proposed?
Physicist Richard Feynman first described the idea of storing information at the molecular level using DNA as a natural model in 1959.
What specific materials comprise the new Penn State memory element?
The technology uses commercially available synthetic DNA doped with silver nanoparticles, layered over a thin film of the semiconductor material perovskite.
Worth a look