Chinese Scientists Invent Chip Similar To A Human Brain

Chinese scientists have invented an advanced chip that is allegedly similar to that of a human brain. The so-called "Darwin" is a joint creation of researchers from both Hangzhou Dianzi University and Zhejiang University.

The human-brain like chip, otherwise called Darwin Neural Processing Unit, is launched earlier this week after a yearlong of hardwork and research. "It can perform intelligent computer tasks by simulating a human brain's neural networks, in which neurons connect with another via synapses," Dr. Ma De of Hangzhou Dianzi University said.

The new technology that can carry out sophisticated tasks and decode ambiguous information is actually only 5 x 5 mm2 in area size. But it is not something to belittle as it is built-in with two of the building blocks of a human brain -- neurons and synapses.

Scientists admit that installing large-scale device on a small chip is difficult. Yet hard work pays off. With its 2048 neurons and 4 million synapses, the new chip when installed in a computer can do way more than the traditional computer chips, which usually are unable to quickly process. It is pieced together using an 180 nm CMO technology. It can consume 0.84 W/MHz and has 1.8 V power supply.

Shen Juncheng, a scientist from Zheijang Univeristy, claimed that among the complex capabilities that chip can do are recognizing written numbers by various individuals, differentiate varied photos, move on-screen items by receiving brain signal from the user and decode electroencephalogram (EEG) results.

Even though it is still in its initial phase of development, this new technology is expected to be utilized in the fields of robotics, intelligent hardware systems and brain-computer interfaces. "It is the dream of computer scientists to design a more electricity-saving, efficient and intelligent computer system by mimicking the structure and mechanism of human neural networks," Zhu Xiaolei, a Zheijang University doctoral student, said.

Join the Discussion

Recommended Stories

Real Time Analytics