Die-to-Die PLL Clock Alignment for Low-Latency High-Bandwidth Links
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing clock signal synchronization techniques in multiple-die systems face challenges with high latency and low bit rate due to asynchronous clock relationships and physical separation between dies, leading to increased power consumption and reduced data throughput.
Innovation Solution
Implementing a synchronized PLL circuit using a global clock signal as a reference, combined with tuned delay circuits to address clock setup and hold mismatch across different clock domains, eliminating the need for additional buffering circuits and mitigating latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional asynchronous clock synchronization techniques are used in multiple-die systems, then die-to-die communication can be established, but latency increases and bit rate decreases
Solution Approach 1:
The patent merges system-synchronous and source-synchronous clock distribution approaches into a hybrid architecture. The system-synchronous PLL provides a common reference clock across all dies to establish a unified time base, while source-synchronous delay circuits are integrated at each die to locally adjust for timing skew. This combination allows data and clock signals to be launched simultaneously from each die while maintaining global synchronization, thereby reducing latency and increasing bit rate compared to traditional asynchronous approaches.
2Reliability
If additional buffering circuits are added to synchronize clock domains, then data transfer reliability improves, but device complexity and power consumption increase
Solution Approach 1:
The patent extracts the synchronization function from separate buffering circuits and integrates it directly into the clock distribution and data launch path. By using PLLs to generate synchronized clock signals and delay circuits to adjust timing at the source, the need for additional buffering circuits is eliminated. This integration maintains data transfer reliability through precise timing control while reducing device complexity and power consumption.
3Quantity of substance
If physical separation between dies is increased for advanced packaging, then integration density improves, but clock synchronization difficulty increases leading to higher latency
Solution Approach 1:
The patent introduces PLL circuits as intermediary devices at each die that generate local clock signals based on a reference clock received from other dies. This intermediary mechanism allows each die to independently generate synchronized clock signals despite physical separation, compensating for propagation delays through local phase adjustment. The delay circuits act as additional intermediaries to fine-tune timing, enabling high integration density through advanced packaging while maintaining low latency through active synchronization.
Data Source
AI summary
A semiconductor device is provided. The semiconductor device includes a first die having a transmission circuit and a first phase-locked loop (PLL) circuit configured to generate a first global clock signal. The semiconductor device includes a second die having a receiver circuit, a phase aligned element, and a second PLL circuit. The phase aligned element is configured to generate a reference clock signal using the first global clock signal and feedback from the second PLL circuit. The second PLL circuit configured to generate a second global clock signal based on the reference clock signal. The phases of the first global clock signal and the second global clock signal are aligned to facilitate data transfer from the transmission circuit to the receiver circuit.


