Semiconductor Memory Data Input Buffer Timing Equalization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

High-speed semiconductor memory apparatuses face operational stability and area efficiency issues due to decreased timing margins between internal clock signals and input data signals, leading to coupling noises and obstructed integration.

Innovation Solution

The semiconductor memory apparatus employs synchronous and asynchronous data input buffers with transmission lines of equal lengths for both clock and data signals, allowing for synchronized buffering operations and improved timing margins, while allowing for non-adjacent buffer arrangements to enhance area efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If data input buffers are densely arranged adjacent each other to reduce timing differences, then timing stability is improved, but occupational area increases and integration is obstructed

Engineering Contradiction:
Improvetiming stabilityVSAvoidoccupational area
Core Design Contradiction:
ReliabilityVSArea of stationary object

Solution Approach 1:

The patent segments the data input circuit into multiple independent data input buffers (first data input buffer, second data input buffer, etc.) that can be arranged in a distributed manner rather than densely packed. Each buffer is independently connected to data pads through separate data lines, allowing spatial distribution that reduces area occupation while maintaining timing stability through equal-length data line design.

Inventive Principle:
Principle #1Segmentation

2Adaptability or versatility

If data lines have different configuration pathways to connect data pads and data input buffers, then routing flexibility is improved, but coupling noises occur and operational stability degrades

Engineering Contradiction:
Improverouting flexibilityVSAvoidoperational stability
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent applies local quality by making each data line have equal length specifically, while allowing different configuration pathways. This localized constraint (equal length) is applied to each individual data line connection between data pads and data input buffers, ensuring that timing characteristics are uniform across different routing paths, thereby preventing coupling noises and maintaining operational stability.

Inventive Principle:
Principle #3Local quality

3Speed

If internal clock frequency is increased for high-speed operation, then processing speed is improved, but timing margin decreases and operational stability degrades

Engineering Contradiction:
Improveprocessing speedVSAvoidtiming margin
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The patent performs preliminary action by pre-configuring equal-length data lines before the actual data input operation. This preliminary design ensures that signal transmission time is equalized across all data paths, allowing the system to operate at high clock frequencies without timing violations, thereby maintaining adequate timing margins even at high speeds.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS7826306B2Semiconductor memory apparatus
Publication Date: 2010.11.02 SK HYNIX INC
  • US7826306B2 patent drawing
  • US7826306B2 patent drawing
  • US7826306B2 patent drawing

AI summary

A semiconductor memory apparatus includes a clock generator configured to generate an internal clock signal, an asynchronous data input buffer configured to buffer a data input signal through a data pad to output a buffered data signal, and a synchronous data input buffer configured to buffer the buffered data signal synchronously with the internal clock signal, wherein a length of a line, through which the internal clock signal is transmitted to the synchronous data input buffer, is configured to be substantially the same with a length of a line, through which the buffered data is transmitted to the synchronous data input buffer.