0
$\begingroup$

Preface: I’d like to clarify that I understand what a relay is and that a PLC uses a fairly conventional microprocessor that only digitally establishes logical logic gate configuration as a digitally programmable alternative to relay banks for analog and/or (depending on the PLC) digital signals. My question is based on the understanding that to date actual logic gates (as far as I know) aren’t non-locally programmable (“re-wirable”) without a person manually rewiring truly programmable actual (not logical programming of a statically wired microprocessor) logic gates.

Rectenna work interests me specifically around any potential relevance of varying transmission wavelengths and material resistances (if this is not possible with MoS2, generally as a concept for other potential materials) to making possible remote switch activation of logically chosen switches along an array. Essentially I am curious about if this or other research has potential for constructing truly physically reprogrammable (externally and maybe wirelessly) logic gates.

In general any information on advances towards this capability would be appreciated as right now it seems like the only rudimentary build I could manage for my project is a 64 gate one. That’s not great because anything less than 512 gates would be very hard to make useful for my proof of concept project, and I know there’s no way I could get to a more ideal 262,144 gates.

One example would be any publication which covers if the kind of uses of phase-engineered low-resistance contacts for ultrathin MoS2 transistors covered in the articles below would be able to be produced with varying resistance in a band usable for varying activation via radio waves for switches.

https://doi.org/10.1038/nmat4080

https://www.ece.cmu.edu/news-and-events/story/2019/05/rectennas-converting-radio-waves-into-electricity.html

I’m not picky if someone knows about other technological advances approaching this capability such as biochemical non-locally programmable switch activation equivalent processes. Thanks everyone.

Update 1: My specific question is: Have there been any significant technological advances towards non-locally electrically programmable logic gates?

Update 2: After further review I’ve found that FPGAs are not what I am asking about. Their reprogramming like PLCs is digital not analog. They seem to just be a more generalized similar thing to PLCs rather than being factory equipment. I might incorporate one or more in my project, but they aren’t what I am referring to which is true analog reprogramming. Why does analog matter? Analog means more efficient at the surface level, but it also allows structured logic similar to ladder logic at the hardware level which enables significantly different uses in structuring and restructuring logic execution.

Update 3: This is for an efficiency proof of concept project trying to prove it is possible to structure logic in a certain way to increase efficiency of certain specific processes. This is a project involving programming and/or design at every single level of development (transistors, machine code, assembly, mid level (such as C/C++), and high level (Python/Tensorflow). I will be creating custom NAND gate structures, writing the instructions to execute on them, writing in assembly, writing in a mid level language, and writing in Python and TensorFlow for different parts of this overall project’s functionality.

In conclusion the straightforward version of this question is: What are the current capabilities for or research done towards creating physically rewired logic gates using non-local digital instructions?

$\endgroup$
6
  • $\begingroup$ Hello. Welcome to AI SE. Can you clarify how is this question on-topic on this site according to ai.stackexchange.com/help/on-topic? So, please, before answering, take the time to read our on-topic page that I just linked you to. $\endgroup$
    – nbro
    Aug 17 at 10:53
  • $\begingroup$ “Questions related to specific software, hardware, or datasets are off-topic, including…“ This would be on topic because it is about developing a machine learning proof of concept proving that the number of operations required to perform a task using in part physically reprogrammable logic gates is possible. Specifically I am trying to train a high level model that can select NAND gate structures to integrate processing into a ladder logic-esque process based on external (non-local to this process) processing inputs and outputs across a few systems. I can clarify further if needed. $\endgroup$
    – Anony Mous
    Aug 17 at 13:46
  • $\begingroup$ Sorry my first comment had a mistake. I meant proving the number of operations required to perform a specific task can be reduced using…. logic gates. (Is possible isn’t part of the properly rephrased sentence) Basically I’m trying to prove that a specific machine learning task can be made more efficient using in part what I’m asking about. $\endgroup$
    – Anony Mous
    Aug 18 at 5:15
  • $\begingroup$ You might be interested to learn about FPGAs. What do you think is the difference between "analog reprogramming" and "digital reprogramming"? You want a robot that physically moves wires around? Why? $\endgroup$
    – user253751
    Aug 18 at 10:29
  • $\begingroup$ I thought I mentioned both PLCs and FPGAs aren’t what I’m referring to. My bad. I want the equivalent of a person moving wires around because there are things you can do with that that you cannot do with programming NAND gates in an FPGA. You can’t integrate data flows based on programmed I/O (think PLC-like outputs) into the structure of an FPGA in a way that suits machine learning the way I plan to do it without it being inefficient which defeats the whole point. You can’t make 3 dimensional gate structures, and you can’t make sets of gate that conditionally are or aren’t connected. $\endgroup$
    – Anony Mous
    Aug 18 at 14:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.