Interposer Capture Buffer Reduction for High-Speed PCIe Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional PCIe Protocol Analyzers face challenges with high costs and error proneness due to the need for large memory buffers to capture and process high-speed data, particularly in the PCIe Gen6 protocol, which requires up to 128 GB/s bandwidth, leading to issues with signal integrity and data processing.

Innovation Solution

The implementation of an interposer circuit with a capture buffer that performs real-time data integrity checks, omits correct data integrity bits, performs ACK/NACK matching, removes or reduces unnecessary fields, and compresses data payloads in parallel to reduce the amount of data stored, using a system with a user interface computer and a high-speed data link for analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If large memory buffers are used to capture high-speed PCIe Gen6 data, then data capture capability is improved, but cost and device complexity increase

Engineering Contradiction:
Improvedata capture capabilityVSAvoidmemory buffer size
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent extracts and removes redundant data elements from the captured PCIe protocol data before storage. Specifically, it removes data integrity bits for correct packets, ACK/NACK packets that are not needed for analysis, and other unnecessary fields, keeping only the essential data needed for protocol analysis. This extraction approach maintains data capture capability while significantly reducing the volume of stored data.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies discarding by eliminating data that does not contribute to analysis value. Correct data integrity bits are discarded after verification, ACK/NACK packets are discarded after matching, and redundant fields are removed. The system recovers only the necessary information needed for protocol debugging and analysis, achieving efficient data reduction.

Inventive Principle:
Principle #34Discarding and recovering

2Quantity of substance

If large memory buffers are used to store captured data, then data storage capacity is improved, but signal integrity deteriorates

Engineering Contradiction:
Improvedata storage capacityVSAvoidsignal integrity
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent extracts only the essential data elements needed for protocol analysis, removing redundant information that would contribute to data volume without adding analytical value. By extracting only necessary packets and fields, the system reduces the total data volume stored, thereby improving signal integrity while maintaining adequate storage capacity for meaningful analysis.

Inventive Principle:
Principle #2Taking out (Extraction)

3Measurement precision

If all data fields are stored for complete analysis, then measurement precision is improved, but data processing complexity increases

Engineering Contradiction:
Improveanalysis completenessVSAvoiddata processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies local quality by treating different data fields differently based on their analytical value. Essential fields such as packet headers, payload data, and error indicators are preserved with full detail, while redundant fields such as correct data integrity bits and routine ACK/NACK packets are removed. This selective preservation maintains measurement precision for critical analysis while reducing overall processing complexity.

Inventive Principle:
Principle #3Local quality

4Quantity of substance

If data is compressed and reduced before storage, then storage requirements are reduced, but information loss may occur

Engineering Contradiction:
Improvestorage requirementsVSAvoiddata information completeness
Core Design Contradiction:
Quantity of substanceVSLoss of information

Solution Approach 1:

The patent uses extraction to remove only redundant information that does not contribute to analysis value, such as correct data integrity bits and unnecessary ACK/NACK packets. This selective removal reduces storage requirements while preserving all information necessary for protocol analysis, avoiding any loss of meaningful data.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent discards data elements that have already served their purpose or are redundant for analysis. Correct packets have their integrity bits discarded after verification, and ACK/NACK packets are discarded after matching. This discarding approach reduces storage requirements without losing information needed for debugging and analysis.

Inventive Principle:
Principle #34Discarding and recovering

Data Source

PatentUS12052098B2Method and system for reducing data stored in capture buffer
Publication Date: 2024.07.30 KEYSIGHT TECHNOLOGIES INC
  • US12052098B2 patent drawing
  • US12052098B2 patent drawing
  • US12052098B2 patent drawing

AI summary

A method is provided for reducing data stored in a capture buffer of an interposer circuit during communication of the data over a data link according to a high-speed, layered packet-based protocol for analysis. The method includes performing data integrity checks of the data in real time, and omitting data integrity bits corresponding to the data integrity checks from transaction layer packets (TLPs) and data link layer packets (DLLPs) of the data when the data integrity checks indicate the data is correct; performing acknowledge and negative acknowledge (ACK/NACK) matching in real time to confirm successful delivery of the TLPs of the data using ACK/NACK packets, where the ACK/NACK packets are omitted from being stored in the capture buffer; removing and/or reducing fields in real time from the TLPs and/or the DLLPs of the data; and compressing data payloads of the TLPs and/or the DLLPs of the data in parallel.