IP Library Granted Patent US 12688406
Granted Patent B2
US 12688406 · App. 17/955,959 · Granted Jul 21, 2026

Charge domain mathematical engine and method

Inventors: David Schie (Houston, TX); Sergey Gaitukevich (Houston, TX); Peter Drabos (Houston, TX); Andres Sibrai (Houston, TX); Erik Sibrai (Houston, TX)
G06N3/065G06G7/16G11C11/54G11C19/0883H03K3/02H03K23/00H10D30/6892H10D30/69G06G7/1907
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12688406
App. No.
17/955,959
Granted
Jul 21, 2026
Kind
B2
Abstract

A multiplier has a pair of charge reservoirs. The pair of charge reservoirs are connected in series. A first charge movement device induces charge movement to or from the pair of charge reservoirs at a same rate. A second charge movement device induces charge movement to or from one of the pair of reservoirs, the rate of charge movement programmed to one of add or remove charges at a rate proportional to the first charge movement device. The first charge movement device loads a first charge into a first of the pair of charge reservoirs during a first cycle. The first charge movement device and the second charge movement device remove charges at a proportional rate from the pair of charge reservoirs during a second cycle until the first of the pair of charge reservoirs is depleted of the first charge. The second charge reservoir thereafter holding the multiplied result.

Claims (29)

1 . A multiplier comprising:

a pair of charge reservoirs, wherein the pair of charge reservoirs are connected in series;

a first charge movement device inducing charge movement to or from the pair of charge reservoirs at a same rate;

a second charge movement device inducing charge movement to or from one of the pair of reservoirs, the rate of charge movement programmed to one of add or remove charges at a rate proportional to the first charge movement device;

wherein a first charge is loaded into a first of the pair of charge reservoirs during a first cycle, the first charge movement device and the second charge movement device removing charges at a proportional rate from the pair of charge reservoirs during a second cycle until the first of the pair of charge reservoirs is depleted of the first charge leaving a charge representing the first charge multiplied by a ratio of charge movement rates on the second charge reservoir.

2 . The multiplier of claim 1 , further comprising a device to stop charge movement.

3 . The multiplier of claim 1 , wherein the first charge movement device and the second charge movement device are based upon control of electric fields.

4 . The multiplier of claim 1 , wherein the first charge movement device is a first current source coupled in series with the pair of charge reservoirs and the second charge movement device is a second current source coupled at a node where the pair of charge reservoirs meet.

5 . The multiplier of claim 2 , wherein the device to stop charge movement is a transfer gate, the transfer gate lowers a barrier and then raises the barrier in conformance with an event while maintaining a field to control charge movement at other times.

6 . The multiplier of claim 1 , wherein said pair of charge reservoirs are capacitors.

7 . The multiplier of claim 1 , wherein said pair of charge reservoirs are floating diffusions.

8 . The multiplier of claim 1 , wherein the first charge is introduced to the first of the pair of charge reservoirs by one of a switched capacitor circuit, an active pixel circuit; or a V/I to charge circuit.

9 . An analog multiplier comprising:

an active pixel comprising at least one of a pinned photodiode or other photodetector, wherein input information to the active pixel is stored on a first input charge reservoir;

a second charge reservoir coupled to the first reservoir;

a transfer gate positioned between the first charge reservoir and the second charge reservoir, wherein a first rate of charge of movement may be controlled by a field coupled to the transfer gate; and

a second charge movement device coupled to the second charge reservoir, wherein a second rate of charge movement may be programmed in proportion to that of said first rate of charge movement;

wherein a first charge is loaded into the first charge reservoir only during a first cycle and the transfer gate and the second charge movement device move charge proportionally during a second cycle until the first charge reservoir is depleted to produce a charge multiplication on the second charge reservoir at the end of the second cycle.

10 . The analog multiplier of claim 9 , comprising junction depleted transfer gates to minimize overlap capacitance.

11 . A multiplier comprising:

a pair of charge reservoirs, wherein each of the pair of charge reservoirs are coupled to a gated charge movement device;

wherein the gated charge movement device is programmed so a rate of charge movement is proportional, the gated charge movement device stopping charge movement once one of the pair of charge reservoirs is depleted.

12 . The multiplier of claim 11 , comprising a device loading multiple weighted inputs into the first of the pair of charge reservoirs.

13 . The multiplier of claim 11 , comprising multiple charge movement rate controlled inputs coupled to the first of the pair of charge reservoirs each individually gated in time by a programming device.

14 . The multiplier of claim 12 , wherein the device is a time weighted crossbar comprising:

a plurality of cross bar conductor lines;

a plurality of gated current sources coupling to the plurality of conductor lines, whereby the plurality of gated current sources are gated according to one of an input voltage, current and a timeframe or just a time; and

a controller circuit coupled to each of the plurality of gated current sources and enabling each of the plurality of gated current sources in conformance with a desired neural network configuration summing multiple weighted inputs.

15 . The multiplier of claim 14 , wherein the time weighted crossbar comprises one of an analog memory, a memristor memory, a floating gate memory, a flash memory or a DNA memory to set a gating time.