AI is locked to multi-layer SOFTWARE and expensive HARDWARE for training & deploying large models for commercial use.
HARD-CORE AI CHIPS use gated circuits to model entire neural networks free from software, and will reduce cost to less than one-tenth.
Validated on a FPGA micro-processor; not ideal, but still improves image recognition speed by a factor of 4682 with a comparible ARM based CPU. Scaling to ASIC will unlock a new world of possibilities.
HC-AI CHIPS act like neurons in the brain, in which a trained AI model gets converted into a fixed circuit structure within the chip.
Our chip receives direct input for blazing fast transmition and processes it along a fixed link in the chip.
No cloud, general purpose CPU & GPU required, besides input devices and directing the target output for desired use.