Serialized SGMII Multiplexing to Cut Pin Count and Power

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current Serial Gigabit Media Independent Interface (SGMII) formats require a large number of pins and significant system power for interfacing between Ethernet switches and physical layer devices, leading to increased die area and routing complexity.

Innovation Solution

The Extra Fast Serial Gigabit Media Independent Interface (XFSGMII) method involves multiplexing multiple encoded serial data streams into a single stream, marking and scrambling them to reduce pin count and power consumption, while maintaining independent port operations and spectral characteristics.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If SGMII format is used to interface between Ethernet switches and physical layer devices, then data transmission capability is provided, but pin count increases to over 300 pins

Engineering Contradiction:
Improvedata transmission capabilityVSAvoidpin count
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent combines multiple SGMII data lanes into a single aggregated data stream using a serializer circuit. Four separate 1.25 Gb/s lanes are merged into one 5 Gb/s lane, reducing the number of physical pins required from over 300 to just a few pins for the aggregated connection.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent transitions from parallel data transmission across multiple lanes to serial transmission over a single aggregated lane. This dimensional change from parallel to serial architecture reduces pin count while maintaining data transmission capability through time-division multiplexing.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If SGMII interface is implemented with multiple pins and buffers, then data transmission is enabled, but system power consumption increases to 7.2 Watts

Engineering Contradiction:
Improvedata transmissionVSAvoidsystem power
Core Design Contradiction:
ProductivityVSUse of energy by stationary object

Solution Approach 1:

The patent merges multiple high-power parallel transmission lanes into a single serialized stream, reducing the number of active I/O circuits and buffers. This consolidation cuts system power consumption from 7.2 Watts to approximately 3.6 Watts while maintaining the same data transmission capability.

Inventive Principle:
Principle #5Merging (Combining)

3Productivity

If multiple pins and buffers are used for SGMII interface, then data transmission capability is achieved, but die area and printed circuit board area increase

Engineering Contradiction:
Improvedata transmission capabilityVSAvoiddie area and PCB area
Core Design Contradiction:
ProductivityVSArea of stationary object

Solution Approach 1:

The patent consolidates multiple I/O circuits and buffers into a single serialized interface, significantly reducing the die area required on the Ethernet switch chip and the PCB area needed for signal routing. The aggregation of lanes eliminates redundant circuitry and reduces routing complexity.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The transition from parallel to serial architecture reduces the spatial requirements for I/O circuits and routing traces. By transmitting data sequentially over a single aggregated lane rather than simultaneously over multiple lanes, the patent minimizes the area occupied by I/O structures and routing paths.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

4Productivity

If SGMII signals are routed over back planes, then connection between devices is established, but expense and routing complexity increase

Engineering Contradiction:
Improvedevice connectionVSAvoidrouting complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent aggregates multiple high-speed signals into a single serialized stream that can be transmitted over a single back plane connection. This consolidation simplifies the routing architecture and reduces the complexity of signal distribution across the back plane while maintaining device connectivity.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentUS7492291B2Methods and apparatus for interfacing a plurality of encoded serial data streams to a serializer/deserializer circuit
Publication Date: 2009.02.17 AVAGO TECHNOLOGIES INTERNATIONAL SALES PTE LTD
  • US7492291B2 patent drawing
  • US7492291B2 patent drawing
  • US7492291B2 patent drawing

AI summary

Methods and apparatus are provided for interfacing a plurality of encoded serial data streams, such as Serial Gigabit Media Independent Interface streams, to a serializer/deserializer circuit. A plurality of encoded serial data streams are transmitted by receiving the plurality of encoded serial data streams that have been encoded using an encoding scheme that provides a substantially uniform distribution of a first code and a second code; marking at least one of the encoded serial data streams (such as changing a first code to a predefined code); and combining at least two of the plurality of encoded serial data streams into a single data stream. A plurality of encoded serial data streams are received by receiving a single data stream comprised of the plurality of encoded serial data streams; detecting a mark in the single data stream; demultiplexing the single data stream into the plurality of encoded serial data streams based on the mark; and providing the demultiplexed plurality of encoded serial data streams to a decoder that decodes the plurality of encoded serial data streams using a decoding scheme that provides a substantially uniform distribution of a first code and a second code.