The human brain is an intricate system. It goes places no one else can and handles multiple tasks that other artificial systems may find difficult. Because of these functions, the human brain has long piqued the interest of scientists, who have conducted numerous studies to better understand it. It is, however, unpredictable. But there’s something else remarkable about the human brain: it uses very little energy. Scientists are attempting to comprehend it in order to draw inspiration for future computer technologies. Experts are concentrating on the brain’s interconnected neuron function because it both stores and processes data.
Scientists anticipate that brain-inspired computers will be more energy-efficient than conventional computers currently in use. They also believe that brain-inspired computers will perform better at machine learning tasks.
The current emphasis on neuron function has permeated the belief that combining storage and processing in a single electronic component type known as a memristor will allow scientists to achieve greater efficiency. Data is moved between the processor and the storage in traditional computers, which is the primary cause of higher energy consumption in machine learning applications.
To accomplish this, the memristor must be modified. Researchers from ETH Zurich, the University of Zurich, and Empa have collaborated to create an innovative concept for memristor that will allow it to be used in a much broader range of applications. In a paper published in the journal Nature Communication, they outline their research.
In a statement, ETH postdoctoral fellow Rohit John explained that memristors have different operation modes. Conventional memristors could only be configured in one of these modes. The new memristors, on the other hand, can easily switch between two operation modes while in use. “These two modes of operation are also found in the human brain,” John explained.
The researchers tested 25 of these new memristors and took 20,000 measurements with them for the study. However, before they can be used in computer technology, these electronic devices will need to be further optimized.
Go home
Visit our YouTube Page
