Samsung is working on ‘Digital Eye’ built on IBM’s TrueNorth processors
Samsung is known for its top notch smartphones including the GALAXY and NOTE series. But, it also works on many other concept products which might never become a reality. The comapany always do research to bring future products. It has already got success in wearable and Virtual reality section. The company is now working on some other projects which might seem to be fiction. It just announced the latest prototype of digital eye that is built on TrueNorth processors which is said to be inspired by IBM’s neurons.
TrueNorth chips, made by IBM, are very different from the regular CPUs found in our day to day electronics. The processor has 4,096 computing cores that are all connected by around 256 million connections. TrueNorth chips works the similar way like the human brain, by each core transmitting messages to each another.
Such neuromorphic chip designs are specially optimized for calculating a very large amount of data, while consuming much lesser power than our regular and simple processors. Samsung had used TrueNorth into its Dynamic Vision Sensor, which analyses each individual camera pixel at a lightning fast speed of 2000 frames per second to capture any changes. This enables the digital eye to accurately track movement in a 3D environment.
Samsung’s digital eye was showcased in the form of a motion gesture controlled TV. Samsung displayed this technology in the form of a TV that could recognize hand gestures and even finger pinch from up to 10 feet of distance. Neuromorphic processors can be used in places like Artificial Intelligence and self-driving cars.
The most eye catching and spectacular thing about Samsung’s digital eye is its energy consumption which is only around 300 milliwatts of power that is almost a one-tenth of your smartphone processor. Following conclusion can be made that this chip or similar designs could be implemented as a separate processor for assistance purpose with today’s CPUs. TrueNorth is the most suitable processors for higher-end systems and Artificial Intelligence research.
It can be clearly seen that there is a lot more research to be done in the field of computer learning and Neuromorphic processors, but we have to agree as to how quickly we are catching up with this emerging field.
For further reading: Beyond von Neumann, Neuromorphic Computing Steadily Advances