If it is medical, it should not be too far away to handle small tasks through simple calculations. Ordinary chip-type ones have been used for a long time. At present, they are relatively practical. Biocomputers require The control is very troublesome, and there are a lot of peripheral devices. It cannot be used in small places now. As for the intelligent application of human thinking, I am afraid it will have to wait until after the human era.
IBM says this chip has the potential to transform the way computers are built in the future, allowing it to consume as much power as a hearing aid battery. Dharmendra Modha, IBM's chief scientist for brain-inspired computing, said the company's long-term goal is to produce a "brain in a box" that can quickly identify patterns in huge data sets using less than 1 kilowatt of power. Various modes. Scientists say that equipment using this chip can be used to screen bioengineering samples to identify harmful factors, provide support for unmanned spacecraft, or monitor abnormal behavior of computer networks.
Modha said that IBM's flagship supercomputer Watson is a representative example of linear computing, and the above-mentioned chip can recognize or "perceive" the surrounding environment in real time, similar to human ears and eyes. For example, the chip could be used to play table tennis, where it can "see" the ball and move the racket to catch it. However, while IBM stressed that this technology may eventually be used in consumer products such as glasses, such applications are still a long way off.
IBM won a $325 million contract from the U.S. Department of Energy last year to develop two advanced supercomputers named "Summit" and "Sierra" by 2017. The two computers may incorporate this new architecture.
Modha said one of the challenges engineers currently face is how to replicate human brains of various sizes and the software used to operate them. IBM started by replicating a human brain the size of a worm, with engineers using 256 processors to simulate neurons, and then scaled up to using 1 million processors, which is equivalent to a human brain the size of a bee. A team of IBM engineers aims to replicate a mouse-sized human brain by the end of next year, using a chip that uses 256 million processors to simulate neurons. But even if this goal is achieved, it is still far from the 100 billion neurons of the human brain.