Voltage Distribution Mixer Circuits for Sub-Threshold Probabilistic Sampling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current integrated circuits (ICs) face challenges in efficiently sampling from a mixture of multiple probability distributions, particularly when operating in the sub-threshold regime, where thermodynamic processes are harnessed for computations.

Innovation Solution

The proposed circuit architecture includes metastable circuits, noise circuits, and mixer circuits to produce and combine voltage distributions, utilizing CMOS transistors in the sub-threshold regime to sample from Gaussian mixture models, employing p-type and n-type metal-oxide-semiconductor transistors and level-shifter circuits to enhance signal processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If current integrated circuits are used for sampling from mixture of probability distributions, then basic computational operations can be performed, but efficiency is insufficient particularly in sub-threshold regime

Engineering Contradiction:

Improvesampling efficiencyVSAvoidcomputational accuracy

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The circuit is divided into specialized functional blocks: metastable circuits for generating probabilistic states, noise circuits for generating voltage distributions, and mixer circuits for combining distributions. This segmentation allows each component to be optimized for its specific function, improving overall sampling efficiency while maintaining accuracy through coordinated operation of specialized subsystems

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The circuit operates in the sub-threshold regime by changing the voltage parameter to be below the transistor threshold voltage. This parameter change enables thermodynamic processes to dominate, allowing the circuit to naturally sample from probability distributions based on Boltzmann statistics, thereby improving sampling efficiency for probabilistic computations

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If metastable circuits operate in sub-threshold regime to harness thermodynamic processes, then computational capabilities are enhanced, but circuit complexity increases

Engineering Contradiction:

Improvecomputational capabilityVSAvoidcircuit architecture

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The mixer circuit serves multiple functions: it combines voltage distributions from different noise circuits, weights them according to probabilistic states from metastable circuits, and outputs a combined distribution. This multi-functionality reduces the need for separate dedicated circuits for each operation, thereby enhancing computational capability without proportionally increasing circuit complexity

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

Noise circuits act as intermediaries between the metastable circuits and the mixer circuits. They convert probabilistic states into voltage distributions that can be systematically combined, providing a standardized interface that simplifies the overall circuit architecture while enabling complex probabilistic computations

Inventive Principle:

Principle #24Intermediary (Mediator)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enables efficient sampling from a mixture of probability distributions, approximating target distributions through Gaussian mixture models, enhancing computational capabilities of ICs in the sub-threshold regime.

Implementation Method 1

a first metastable circuit configured to produce a bistable state that varies over time between a first stable voltage and a second stable voltage, where a fraction of time that the bistable state spends at the first stable voltage is associated with a first probability

Methodology Applied

Scientific EffectMetastability: Metastability

Implementation Method 2

a first noise circuit configured to produce a first voltage distribution; a second noise circuit configured to produce a second voltage distribution

Methodology Applied

Scientific EffectThermal noise:

Implementation Method 3

a mixer circuit configured to receive the first voltage distribution, the second voltage distribution, and one or more signals based at least in part on the bistable state; wherein the mixer circuit is configured to produce a third voltage distribution that is based at least in part on the one or more signals associated with the bistable state, the first voltage distribution, and the second voltage distribution

Methodology Applied

Scientific EffectProbabilistic mixing:

Data Source

PatentEP4687288A1Circuits for mixing voltage distributions associated with random variable samples

Publication Date: 2026.02.04 EXTROPIC CORP

Figure 1A
Figure 1B
Figure 2A

AI summary

A method comprises: producing, using a first metastable circuit, a bistable state that varies over time between a first stable voltage and a second stable voltage, where a fraction of time that the bistable state spends at the first stable voltage is associated with a first probability; producing, using a first noise circuit, a first voltage distribution; producing, using a second noise circuit, a second voltage distribution; and producing, using a first mixer circuit, a third voltage distribution that is based at least in part on the bistable state, the first voltage distribution, and the second voltage distribution.