Scientists have created a new type of memory that will significantly speed up artificial intelligence
A research team led by Cambridge University has developed a new computer memory design that will significantly improve performance and reduce the power consumption of internet and communications technology.
Here's What We Know
According to the university, artificial intelligence, the internet and other data-driven technologies will consume more than 30% of the world's electricity over the next decade. Moving information between storage and processing devices requires a lot of energy and time, the study says.
Scientists experimented with a new type of technology known as resistive switching memory. Unlike existing devices, which encode data in two states (one or zero), the scientists' development allows for a continuous range of states.
This has been achieved by applying an electric current to certain materials, which causes the electrical resistance to increase or decrease. Different changes in electrical resistance create different possible states for data storage.
"A typical USB stick based on continuous range would be able to hold between ten and 100 times more information, for example," said lead study author Markus Hellenbrand.
The team has developed a prototype hafnium oxide-based device. Until now, it proved difficult to use in resistive switching memory because of the material's lack of structure at the atomic level. However, scientists found a solution: they added barium to the mixture.
This created highly structured barium "bridges" between thick hafnium oxide films. At the point where these "bridges" cross the contacts of the devices, an energy barrier is created allowing electrons to cross.
The energy barrier can be raised or lowered. This changes the resistance of the hafnium oxide composite and allows the material to exist in several states.
According to the scientists, the final result was similar to the way synapses in the brain work, which can store and process information in the same place. Researchers believe this could lead to computer memory devices with much greater density and performance, but with lower power consumption.
Cambridge Enterprise, the university's commercial arm, has applied for a patent. The scientists are now working with industry to carry out more extensive research. They claim that integrating hafnium oxide into existing manufacturing processes would not be difficult, as the material is already being used in the manufacture of semiconductors.
Source: The Next Web.
