HYBRID NEURON

HYBRID NEURON COMPUTATION


Neuromorphic Computation is a concept developed by Carver Mead, in the late 1980s, describing the use of very-large-scale integration (VLSI) hardware systems containing electronic analog circuits to mimic neuro-biological architectures present in the nervous system. This is the QAI company focus.

The company objective is high precision EHF analog neural ultra computation and uW wireless integrated circuit self-organizing interconnect. We believe this is the path forward for Artificial Intelligence. There are two reasons for this:

First, digital computation is inherently restricted to rationale number computation due to the Church-Turing Thesis, which greatly limits computational power in comparison to biological brain neural analog computation. Consequently, analog computation transcends digital computation because analog computations includes real number values as proved by the Sontag-Siegelmann hypothesis.

Second, current VLSI wired interconnect strategies provided very limited fan in and fan out compared to a single brain neuron. In this context we pursue superconducting (resistance free) wireless uW Multi-chip module interconnect, which offers the potential of high connectivity compared to current methods.

Our current project is applied to cryptanalysis, implementing a new GaAs/GaN 3THz 4D Multi-Chip Module on micro-cooled diamond/Ni dielectric. Each analog integrated circuit has 1000 analog devices implemented using a low power ultra high frequency HEMT/Schottky diode process. The SLUICE BOX analog recurrent processing performance is three Quadrillion analog computations per second per integrated circuit die.