Gated Computing Memory Cells for Parallel Bit-Line Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The exponential growth of data generation outpaces data processing capabilities, leading to high time and power costs associated with accessing and processing data, as existing technologies struggle to efficiently access and compute large amounts of data simultaneously.
Innovation Solution
A memory device with an array of gated computing cells that allows simultaneous access to multiple memory locations and performs computations in-memory, using logic elements like NOR, OR, AND, and NAND gates to generate signals indicative of data values, reducing the need for external processing and minimizing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is accessed and processed using external processors, then data processing capability is improved, but time cost and power consumption increase
Solution Approach 1:
The patent merges memory storage and processing functions into a single integrated structure. Memory cells store data while logic elements perform computations directly on stored data, eliminating the need to transfer data between separate memory and processing units. This combination resolves the contradiction by improving productivity through in-memory computing while reducing time cost by eliminating data movement overhead.
Solution Approach 2:
The patent introduces bit lines as intermediaries that enable simultaneous read operations across multiple memory cells. These bit lines serve as mediators that collect signals from multiple memory locations in parallel, allowing the system to process large amounts of data simultaneously without increasing time cost, thus resolving the contradiction between productivity and time loss.
2Productivity
If data is accessed and processed using external processors, then data processing capability is improved, but power consumption increases
Solution Approach 1:
By merging memory and processing functions into an integrated architecture where logic elements are embedded within memory cells, the patent eliminates power-consuming data transfers between separate components. Processing operations are performed directly where data is stored, improving productivity while minimizing power consumption by removing the energy overhead of data movement.
Solution Approach 2:
The memory system performs processing operations on its own stored data without requiring external processors. The logic elements within each memory cell enable the memory to service its own computational needs, improving productivity while reducing power consumption by eliminating the energy cost of external data transfer and processing.
3Speed
If simultaneous access to multiple memory locations is performed, then data access speed is improved, but signal equalization complexity increases
Solution Approach 1:
The patent applies partial action by performing computations on a subset of data values directly during the read operation. Logic elements evaluate Boolean functions on selected memory cell outputs without requiring complete signal equalization across all bit lines. This approach maintains high data access speed while reducing equalization complexity by processing only the necessary portion of data.
Solution Approach 2:
The patent replaces the mechanical/electrical signal equalization process with logical evaluation. Instead of physically equalizing signals across bit lines through complex circuitry, logic elements directly evaluate Boolean functions on memory cell outputs. This substitution maintains high access speed while dramatically reducing equalization complexity by using logical operations rather than signal conditioning.
4Productivity
If logic elements are integrated into memory cells, then processing efficiency is improved, but device complexity increases
Solution Approach 1:
The patent implements multi-functionality by designing memory cells that simultaneously perform storage and processing functions. Each memory cell contains both storage elements and logic elements that can evaluate Boolean functions. This universal design improves processing efficiency by enabling in-cell computations while managing device complexity through functional integration rather than adding separate processing components.
Solution Approach 2:
The patent applies segmentation by dividing the processing function into discrete logic elements that are distributed across multiple memory cells. Each logic element handles specific Boolean evaluations on its associated memory cell outputs. This segmentation improves processing efficiency through parallel distributed computation while managing device complexity by breaking down the processing function into manageable modular units.
Data Source
AI summary
Systems and methods are provided for a computing-in memory circuit that includes a bit line and a plurality of computing cells connected to the bit line. Each of the plurality of computing cells includes a memory element, having a data output terminal; a logic element, having a first input terminal, a second input terminal and an output terminal, wherein the first input terminal is coupled to the data output terminal of the memory element, the second input terminal receives a select signal; and a capacitor, having a first terminal and a second terminal, where the first terminal is coupled to the output terminal of the logic element, the second terminal is coupled to the bit line. A voltage of the bit line is driven by the plurality of computing cells.


