Resistive Network Parallel Branches Analog Computing Linearity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Analog computing systems face limitations in dynamic range and linearity due to the non-linear nature of resistive elements, particularly in neural networks where precise control of conductance levels is required for effective signal processing.
Innovation Solution
The implementation of a network of resistive elements with multiple switched resistive branches connected in parallel, allowing for programmable conductance levels and incremental changes, which enhances dynamic range and linearity by limiting selection to higher resistance values and using additional elements in parallel for fine-tuning low-resistance states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single resistive element is used for conductance control, then the device structure is simple, but the dynamic range and linearity are limited
Solution Approach 1:
The single resistive element is segmented into multiple parallel resistive branches (first, second, and third branches with different resistance values). Each branch can be independently switched to provide different conductance levels, thereby expanding the dynamic range while maintaining a manageable device structure through modular design.
Solution Approach 2:
The resistive branches are made dynamically switchable through control signals that can selectively activate or deactivate specific branches. This dynamic configuration allows the system to adapt its conductance level in real-time, providing a wide dynamic range without requiring a completely different device structure for each conductance state.
2Manufacturing precision
If resistive elements operate across full resistance range, then the conductance control is flexible, but the linearity deteriorates due to non-linear nature
Solution Approach 1:
Different resistive branches are designed with specific resistance values optimized for particular conductance ranges. By selecting appropriate branches for specific operating conditions, the system achieves linear conductance control in each local range while maintaining overall flexibility through the combination of multiple branches with different characteristics.
3Measurement precision
If multiple resistive branches are used in parallel, then the programmable weight levels increase, but the device complexity increases
Solution Approach 1:
Multiple resistive branches are merged in parallel configuration within a unified device structure, sharing common control mechanisms and interconnection pathways. This combining approach enables the system to achieve high programmable weight levels through the additive effect of parallel branches while avoiding the complexity multiplication that would result from completely separate control circuits for each branch.
4Productivity
If conventional resistive elements are used, then the device is easy to manufacture, but the write time is long and precision is limited
Solution Approach 1:
The resistive branches are pre-configured with specific resistance values during manufacturing, and switchable connections are pre-established. This preliminary preparation allows the system to achieve fast write times by simply switching between pre-configured states rather than dynamically adjusting resistance values, while maintaining manufacturing simplicity through standard fabrication processes.
Data Source
AI summary
A resistive network include multiple resistive units; each resistive unit is made up of multiple resistive elements, which can be arranged in a parallel configuration. Each of the resistive elements can be programmable (e.g., switched on or off, or set to one of multiple resistance values). Furthermore, a method of analog computing includes configuring multiple resistive elements in each of multiple resistive units and configuring the resistive units into a network. The configuration of the resistive elements can be, for example, arranging them into a parallel combination. The method further includes programming each resistive unit, for example, by switching individual resistive elements into, or out of, the parallel combination.


