On-Chip SerDes for Dual-Protocol PHY via Passive Channel
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current dual-standard physical layer (PHY) interfaces in mobile devices require significant die space and power due to the need for flip-flop pipelining to align skewed data streams from remote DP controllers, leading to inefficiencies in data transmission and increased latency.
Innovation Solution
An on-chip serializer/deserializer (SerDes) architecture is implemented using a passive transmission channel to serialize and deserialize data streams between DP and USB controllers, eliminating the need for flip-flop pipelining and reducing power consumption by allowing direct data propagation between controllers and the PHY interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If flip-flop pipelining is used to align skewed data streams from remote DP controllers, then data transmission reliability is improved, but die space consumption and power consumption increase significantly
Solution Approach 1:
The patent extracts the essential function of data alignment from the complex flip-flop pipelining structure and implements it through a simplified timing recovery mechanism using delay-locked loops (DLL) and buffer adjustment. This removes the bulky flip-flop infrastructure while retaining the core functionality of skew compensation.
Solution Approach 2:
The patent changes the approach from static flip-flop-based alignment to dynamic parameter adjustment using DLL-controlled timing recovery. The system adjusts buffer read/write pointers and delay elements based on detected skew parameters, enabling adaptive alignment without fixed pipelining structures.
2Reliability
If flip-flop pipelining is used to align skewed data streams, then data transmission reliability is improved, but power consumption increases significantly
Solution Approach 1:
The patent removes the power-hungry flip-flop pipelining infrastructure and replaces it with a low-power timing recovery mechanism using DLLs and programmable buffers. This extraction eliminates unnecessary switching activity while maintaining data alignment functionality.
Solution Approach 2:
The system implements self-adjusting timing recovery where the DLL automatically detects and compensates for skew without requiring high-power external control logic. The buffers self-regulate their read/write timing based on incoming data characteristics, reducing overall power consumption.
3Reliability
If flip-flop pipelining is used for data alignment, then skew compensation is achieved, but transmission latency increases
Solution Approach 1:
The patent transitions from fixed-stage flip-flop delay to continuously adjustable buffer timing controlled by DLLs. This allows dynamic optimization of delay parameters to match actual skew measurements, minimizing latency while achieving complete skew compensation.
Solution Approach 2:
The system employs dynamic timing adjustment where buffer pointers and delay elements are continuously adapted based on real-time skew detection. This replaces the static, over-engineered flip-flop pipeline with a flexible, demand-based timing mechanism that introduces only necessary delay.
4Productivity
If multiple controllers are placed adjacent to the PHY interface input port, then data transmission efficiency is improved, but device complexity increases due to multiplexing requirements
Solution Approach 1:
The patent implements a universal PHY interface that can directly accept data streams from multiple controllers (DP, USB3, etc.) without requiring protocol-specific multiplexing logic. The interface is designed to handle different protocols simultaneously through standardized timing recovery and deserialization mechanisms, reducing overall system complexity.
Solution Approach 2:
The patent segments the data transmission path into independent protocol-handling modules that feed into a common PHY interface. Each controller maintains its own data path to the PHY, eliminating the need for complex centralized multiplexing while allowing efficient parallel operation of multiple protocols.
Data Source
AI summary
An on-chip passive transmission channel is provided for the propagation of serialized data from a first controller to a dual-protocol physical layer interface. A second controller for the dual-protocol physical layer interface is located closer on a semiconductor die to the dual-protocol physical layer interface than the first controller.


