Control Register ECC Scrubbing for Lockstep Computing Reliability

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing computing systems with high reliability requirements, such as those in vehicles, face challenges in maintaining robustness due to the need for redundant processing resources, particularly in managing control registers which are not efficiently protected against data corruption.

Innovation Solution

Implementing end-to-end error correction code (ECC) protection for control registers outside the lockstep unit, combined with periodic scrubbing to validate data integrity, while using redundancy for status registers to enhance system robustness without increasing footprint.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If redundant control registers are implemented to protect against data corruption, then system reliability is improved, but device complexity and resource footprint increase

Engineering Contradiction:
Improvesystem reliabilityVSAvoiddevice complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

An ECC unit is introduced as an intermediary component that generates and validates error correction codes for control register data. This mediator enables reliable error detection and correction without requiring redundant control registers, thus improving reliability while avoiding the complexity of full redundancy implementation.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

Instead of copying entire control registers to create redundant instances, the patent creates a simplified copy in the form of error correction codes (ECC). These ECC bits are stored alongside the original data and used to detect and correct errors, providing protection without the resource overhead of full register duplication.

Inventive Principle:
Principle #26Copying

2Reliability

If redundant processing resources are implemented to increase robustness, then system reliability is improved, but quantity of resources increases

Engineering Contradiction:
ImproverobustnessVSAvoidquantity of resources
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent applies different protection strategies to different types of registers based on their specific needs. Control registers receive ECC protection which is more resource-efficient, while status registers continue to use traditional lockstep redundancy. This localized approach optimizes resource usage by applying the most efficient protection method to each register type.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent changes the protection parameter from full register duplication to bitwise error correction codes. By transforming the protection mechanism from a structural redundancy (copying entire registers) to a data-level redundancy (ECC bits), the system achieves robustness with significantly reduced resource consumption.

Inventive Principle:
Principle #35Parameter changes

3Manufacturing precision

If control registers are protected with ECC, then data integrity is improved, but manufacturing complexity increases

Engineering Contradiction:
Improvedata integrityVSAvoidmanufacturing complexity
Core Design Contradiction:
Manufacturing precisionVSEase of manufacture

Solution Approach 1:

The ECC codes are generated and embedded in the control register data during the write operation, before the data is stored. This preliminary action ensures that error correction capability is built into the data structure at the time of creation, simplifying the manufacturing process compared to post-manufacturing verification methods.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS12499008B2Control register related computing system and method
Publication Date: 2025.12.16 INFINEON TECHNOLOGIES AMERICAS CORP
  • US12499008B2 patent drawing
  • US12499008B2 patent drawing
  • US12499008B2 patent drawing

AI summary

One or more computing devices, systems, and/or methods are provided. In an example, a computing system includes a first processing complex including a first status register unit, a second processing complex including a second status register unit, a lockstep unit configured to verify redundancy in the first status register unit and the second status register unit, a control register unit connected to the lockstep unit and including control registers, each configured to store control register data and an error correction code (ECC) associated with the control register data, a scrubber configured to read the control registers at a predetermined frequency, and an ECC unit configured to, responsive to a read of a selected control register of the control registers by the scrubber, validate the control register data associated with the selected control register based on the ECC associated with the control register data.