• New Hardware NN Approach Using Standard CMOS Transistors

    From c186282@21:1/5 to All on Tue Apr 15 21:02:59 2025
    https://scitechdaily.com/ai-breakthrough-scientists-transform-everyday-transistor-into-an-artificial-neuron/

    The NUS research team has now demonstrated that a single,
    standard silicon transistor, when arranged and operated in
    a specific way, can replicate both neural firing and synaptic
    weight changes — the fundamental mechanisms of biological
    neurons and synapses. This was achieved through adjusting
    the resistance of the bulk terminal to specific values,
    which allow controlling two physical phenomena taking place
    into the transistor: punch through impact ionization and
    charge trapping. Moreover, the team built a two-transistor
    cell capable of operating either in neuron or synaptic
    regime, which the researchers have called “Neuro-Synaptic
    Random Access Memory”, or NS-RAM.

    “Other approaches require complex transistor arrays or novel
    materials with uncertain manufacturability, but our method
    makes use of commercial CMOS"

    . . .

    They CLAIM low power consumption.

    Anyway, if adaptable to the larger scale, this can
    be a useful new way to do hardware NNs.

    Of course we WILL need Linux apps to help train
    these things .......

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From rbowman@21:1/5 to All on Wed Apr 16 06:05:45 2025
    On Tue, 15 Apr 2025 21:02:59 -0400, c186282 wrote:


    “Other approaches require complex transistor arrays or novel materials
    with uncertain manufacturability, but our method makes use of commercial CMOS"

    Yeah, but is it as kool as the Mk. I Perceptron?

    https://en.wikipedia.org/wiki/Mark_I_Perceptron

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)