Switching Buffer Unit for Memory Data Path Routing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In semiconductor memory devices, increasing data transfer speed to achieve higher memory access rates often results in larger chip sizes due to the need for more data buses, creating a trade-off between speed and chip size reduction.

Innovation Solution

The implementation of a switching buffer unit that controls propagation delays by configuring data path routes between memory banks and data input/output terminals based on bank addresses and command signals, minimizing differences in data propagation delay and optimizing data access time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If the number of data buses is increased to achieve faster data transfer speed, then memory access rate is improved, but chip size becomes larger

Engineering Contradiction:
Improvedata transfer speedVSAvoidchip size
Core Design Contradiction:
SpeedVSArea of stationary object

Solution Approach 1:

The patent merges multiple data bus functions into a single shared data bus by implementing a switching buffer unit that dynamically routes data between multiple memory banks and the single data bus. This consolidation reduces the total number of data buses from multiple separate buses to one shared bus, thereby reducing chip size while maintaining high data transfer capabilities through time-division multiplexing.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces a dynamic switching buffer unit that can change data path routes based on bank addresses and command signals. This dynamic routing capability allows the system to adaptively allocate the single data bus to different memory banks as needed, enabling high-speed access to multiple banks sequentially without requiring dedicated buses for each bank, thus resolving the contradiction between speed and chip size.

Inventive Principle:
Principle #15Dynamics

2Reliability

If multiple data buses are aligned with intervals in chip layout to avoid noise effects, then signal integrity is improved, but space in chip layout increases

Engineering Contradiction:
Improvesignal integrityVSAvoidchip layout space
Core Design Contradiction:
ReliabilityVSArea of stationary object

Solution Approach 1:

The patent consolidates multiple separate data buses into a single shared data bus, eliminating the need for multiple spaced-apart bus lines in the chip layout. This merging approach reduces the total layout space required while maintaining signal integrity through the use of switching buffers that isolate and control signal flow, preventing noise interference between different memory bank operations.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The switching buffer unit acts as an intermediary between multiple memory banks and the single data bus. It provides isolation and controlled signal routing, ensuring that data from different memory banks does not cause noise interference on the shared bus. This mediator approach maintains signal integrity without requiring multiple physically separated buses, thus reducing chip layout space.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS10725943B2Apparatuses and methods for transferring data from memory on a data path
Publication Date: 2020.07.28 MICRON TECHNOLOGY INC
  • US10725943B2 patent drawing
  • US10725943B2 patent drawing
  • US10725943B2 patent drawing

AI summary

Apparatuses and methods for transferring data from memory on a data path are described. An example apparatus includes: one or more data terminals; a plurality of memory banks, one of the plurality of memory banks being selected responsive, at least in part, to a bank address; and a data path including a plurality of data path routes and a plurality of switching buffers on the plurality of data path routes. The plurality of switching buffers are arranged such that one or more of the plurality of switching buffers are selected responsive, at least in part, to the bank address and activates one of the plurality of data path routes.