A new generation of computer chips modeled after the brain’s neural network could be launched before the end of this decade thanks to a new material that has been developed.
In a nutshell, it is the first electrochemical 3-terminal transistor manufactured with 2D materials.
Scientists from KTH Royal Institute of Technology in Stockholm and from Stanford University revealed that memory components made with a titanium carbide compound called MXene exhibited “outstanding potential for complementing classical transistor technology”.
Brain-esque PC
The electrochemical random access memory or ECRAM behaves as a synaptic cell in an artificial network, providing a one-stop-shop for storing and processing data. “These new computers [would] rely on components that can have multiple states, and perform in-memory computation”, KTH Associate Professor and lead author Max Hamedi said in a statement.
The findings, published in the journal “Advanced Functional Materials”, suggest that the MXenes could play a fundamental role in developing neuromorphic computers that are closer in operation to human brains and thousands times more energy efficient than today’s traditional computers.
“One million times more efficient”
In a statement to TechRadar Pro, Max Hamedi confirmed that the technology “uses the same processes as cmos wafer assembly, integrating layers of 2D material on silicon, so it’s a true hybrid integration with the same back of the line processes”.
He added, “we see write speeds that are 1000 some faster than any other ECRAM that has been shown. That means that if we scale 2D ECRAMs to nano dimension, they can be as fast as the transistors in today’s computer (sub nanosecond), which means it can fuse into our current computers using CMOS technology process (thanks to the compatibility of 2D transistor metal materials with CMOS fab process).
“We will [therefore] be able to fabricate special purpose computer blocks (in say 5-10 years) where memory and transistors merge making them at least 1000 times more energy efficient than the best computers we have today for AI and simulation tasks (some calculation even show 1 million fold energy efficiency for certain algorithms).
We can likely expect the first commercial product to land before the end of this decade as the GTM (Go to Market) strategy requires at least five years of trials.
