Are ASIC chips the equivalent of "State Machines" in computer science ? There are no "programs" or "instructions" that are executed. All the changes occure because of the inputs on the gates when the clock changes state. Everything happens because of transitions in logic blocks or gates as controlled by the input clock signal.
Are there ASICs that are not clock limited? They are just TTL/MOSFET logic that "run free"? Once the input changes, the ASIC chip free wheels until reaching a stable state.
Why ask? State machines taught in Computer science are implemented with logic gates. State diagrams help form the steps or actions of the state machine like a programming language controls defines the steps taken. Then a compiler turns teh program into machine instructions. I hoped a state machine once defined could then be turned into an ASIC. (Yes FPGA testing and a hundred other steps done as well)
ASICs look to be physical incarnation of the desired state machine written in the I.C. Problem to be solved: faster program execution. Should I use CUDA for massivly parallel execution, ASICs or build a state machine out of MOSFET parts. Quantum computing and a CRAY supercomputer is beyond my reach.
As dumb as my question is, I have tried to phrase it the best way I know. The goal? ASIC implimented AI. Yes ASICs can not be "re programmed", it seems AI could still be implemented -like a neural network. Why? Because our brains do not loose connections once formed. Our brains are very hard to re program but they do learn and self correct.
May you enjoy the chuckle. Sorry for wasting anyone's time. Thanks for any help (other than "run along and play on the highway at night)