It’s known as mimetic technology, machinery that mimics the function and behavior of organic life. For some time, scientists have been using this philosophy to further develop computing, a process which many believe to be paradoxical. In gleaming inspiration from the organic world to design better computers, scientists are basically creating the machinery that could lead to better organics.
But when it comes to Neuromoprhic processors, computers that mimic the function of the human brain, scientists have been lagging behind sequential computing. For instance, IBM announced this past November that its Blue Gene/Q Sequoia supercomputer could clock 16 quadrillion calculations per second, and could crudely simulate more than 530 billion neurons – roughly five times that of a human brain. However, doing this required 8 megawatts of power, enough to power 1600 homes.
However, Kwabena Boahen, a bioengineering professor at Stanford University recently developed a new computing platform that he calls the “Neurogrid”. Each Neurogrid board, running at only 5 watts, can simulate detailed neuronal activity of one million neurons — and it can now do it in real time. Giving the processing to cost ratio in electricity, this means that his new chip is roughly 100,000 times more efficient than other supercomputer.
What’s more, its likely to mean the wide-scale adoption of processors that mimic human neuronal behavior over traditional computer chips. Whereas sequential computing relies on simulated ion-channels to create software-generated “neurons”, the neuromorphic approach involves the flow of ions through channels in a way that emulates the flow of electrons through transistors. Basically, the difference in emulation is a difference between software that mimics the behavior, and hardware.
What’s more, its likely to be a major stepping stone towards the creation of AI and MMI. That’s Artificial Intelligence and Man-Machine Interface for those who don’t speak geek. With computer chips imitating human brains and achieving a measure of intelligence which can be measured in terms of neurons and connections, the likelihood that they will be able to merge with a person’s brain, and thus augment their intelligence, becomes that much more likely.