A computer is a great tool for statistical analysis, simulation...

A computer is a great tool for statistical analysis, simulation and number crunching, but its usefulness is limited in cognitive processing, particularly that entails the simulation of biologically accurate neural networks. This is due to the sequential nature of computers, whereby all data has to pass through a central processor in chunks of 16, 32 or 64 bits. In contrast, the brain's network is massively parallel and processes the equivalent of millions of data bits simultaneously. Besides this advantage, the brain's architecture is such that no obstructions exists in the stream of information. Simulation of a sizable network of biologically accurate neurons has up to now required the resources of a huge supercomputer consisting of hundreds of thousands, even millions of processors. Even so, attempts to emulate the entire human brain are far removed from their goal. The published results of IBM's efforts to emulate the brain on a Blue Gene/Q (Sequoia) Supercomputer, has shown that this machine, based on 1.5 million PowerPC processors runs the emulation of the equivalent of a monkey brain at 1/1542 th (0.065%) of it real-time speed. However, doubts have been casts upon the accuracy of the neuron model that is used in these simulations i . Particularly the synapse function, the release of neurotransmitters, peptides and neuro-modulators are ignored. Presented here is a new digital technology that overcomes these concerns. BrainChip is a technology platform that has the same architecture as the brain and it provides the same parallel processing capability together with its learning abilities. It runs at the same speed as the brain regardless of the size of the emulation. BrainChip consists of an RTL library that contains the tools to construct any type of brain structure to a high degree of accuracy in hardware, rather than software.