Neuromorphic computing is made possible by building "chips" that embody neural networks rather than simply emulating them with traditional Von Neumann chip architectures.
Neuromorphic chips will alter the economics of computing, the energy consumption of computing and the performance equations for computing.
It will find applications in cognitive computing, analog sensing, data conversion, analog and digital signal processing, programmable logic, and associative memory. Deep learning capabiities will be extended and driven closer the periphery of the internet in both mobile devices and the internet of things. Human machine interfaces will be deeply enhanced as neuromorphic computing is rolled out.
It is probable that neuromorphic chips and Von Neumann chips will work in tandem to use the strengths of each.
