Better computers thanks to perovskite nanocrystals
Researchers at Empa, ETH Zurich and the Politecnico di Milano are developing a new type of computer component that is more powerful and easier to manufacture than its predecessors. What is special about it is that it is designed to process large amounts of data quickly and energy-efficiently, following the example of the human brain.
The human brain is still superior to modern computers in many respects. While most people can't do math as well as a computer, we can effortlessly process complex sensory information and learn from our experiences, something a computer can't do (yet). And in the process, the brain consumes barely half as much energy as a laptop.
One of the reasons for the brain's energy efficiency is its structure. The individual neurons and their connections, known as synapses, can both store and process information. In computers, on the other hand, the memory is separate from the processor, and data must be transported back and forth between these two units. The speed of this transport is limited, which makes the whole computer slower when the amount of data is very large.
One possible solution to this bottleneck is novel computer architectures modeled on the human brain. To this end, scientists are working on so-called memristors: Components that, like brain cells, combine data storage and processing. A team of researchers from Empa, ETH Zurich and the Politecnico di Milano has now developed a memristor that is more powerful and easier to manufacture than its predecessors. The researchers recently published their results in the journal "Science Advances".
Performance thanks to mixed conductivity
The novel memristors are based on halide perovskite nanocrystals, a semiconductor material known from the production of solar cells. "Halide perovskites conduct both ions and electrons," explains Rohit John, who until recently was an "ETH Fellow" and postdoc at ETH Zurich and Empa. "This dual conductivity enables more complex computations that more closely approximate brain processes."
The researchers conducted the experimental part of the study entirely at Empa: They fabricated the thin-film memristors in the Thin Films and Photovoltaics Laboratory and investigated their physical properties in the Transport at Nanoscale Interfaces Laboratory. Based on the measurement results, they then simulated a complex computational task that corresponds to a learning process in the visual cortex of the brain. The task was to determine the orientation of a light bar based on signals from the retina.
"To our knowledge, this is only the second time this type of calculation has been performed on memristors," says Maksym Kovalenko, ETH professor and head of the Functional Inorganic Materials research group at Empa and ETH Zurich. "At the same time, our memristors are much easier to manufacture than previous ones." This is because, unlike many other semiconductors, perovskites do not require high temperatures for crystallization. In addition, the new memristors do not require the complex preconditioning by specific electrical voltages that comparable components need for such computing tasks. This makes them faster and more energy-efficient.
Complement, not replace
The technology is not quite ready for use yet. At the same time, the simplicity of manufacturing the new memristors makes them difficult to integrate with existing computer chips: perovskites cannot withstand the temperatures of 400-500 degrees Celsius needed to process silicon - at least not yet. But according to Daniele Ielmini, a professor at the "Politecnico di Milano," this integration is the key to success for the new brain-like computer technologies. "Our goal is not to replace classical computer architecture," he explains. "Rather, we want to develop alternative architectures that can perform certain tasks faster and more energy efficiently. This includes, for example, the parallel processing of large amounts of data, which now occurs everywhere from agriculture to space exploration."
Promisingly, there are other materials with similar properties that could be considered for the production of high-performance memristors. "We can now test our memristor design with different materials," says Alessandro Milozzi, a doctoral student at "Politecnico di Milano." "Possibly some of them are more suitable for integration with silicon."
(Source: Empa)
