IBM has announced the first challenge for its future TrueNorth computer chips, so-called "neuromorphic" computers that are physically structured like a brain: to study the mass of data collected by real human brains and make sense of them. These kinds of tasks are nearly impossible for most traditional computers, but IBM hopes that with its new technology it will be able to observe the human brain in real time - and possibly intervene.
Especially large and very dangerous epileptic seizures are the result of large-scale storms of electrical activity that rage in the two hemispheres of the brain. Doctors struggle to find reliable signatures for these storms, so early detection and effective countermeasures are often not possible. But neural networks are perfect for finding such complex models. IBM is feeding its new TrueNorth neuromorphic chips with electroencephalogram (EEG) data from epileptics in the hopes that they can identify patterns associated with major seizures. However, the real potential lies in the applications and unique benefits of neuromorphic architecture.
IBM has already linked conventional digital computers into simulations of the human brain once. It was just as complex, but it worked 1500 times slower than the original. IBM researchers noted that the real limiting factor was not computing power, but electricity. If we take the conventional digital algorithms used to run a slow simulation and just speed them up 1500 times, we need about 12 gigawatts of power. And until laboratories are equipped with nuclear reactors, a better way will have to be found.
Fortunately, we have a good blueprint: the human brain. After all, the human brain operates with all the complexity of the human brain all the time and requires only 20 watts - enough to power a small light bulb. The IBM TrueMorph neuromorphic chips are also more energy efficient.
These chips are perfect for running code that analyzes data. Sophisticated and subtle data analysis using traditional hardware will discharge a huge battery in a couple of hours; with TrueNorth or a similar architecture, you might be able to wear them on your body and keep them going for a week or even a week.
IBM's ultimate goal is to use TrueNorth to detect seizures in real time - either through an implant or a wearable device. The energy efficient chip will be able to observe and read the user's EEG, identifying an impending seizure. Then he will be able to send you a signal through a smartphone or directly to the first-aid post, and in the future even enter the necessary medications.
Promotional video:
Wearable technologies require a revolution in either energy or energy efficiency. If we are ever going to have devices that get electricity from your clothes, for example, we need chips that do not require large amounts of power. Neuromorphic chips may be one such solution. Monitoring epilepsy is just one possible application.
