Order-Independent Write Signatures for Cross-Machine Error Checking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data processing devices face challenges in detecting errors caused by ionizing radiation and hardware issues, as error detection codes may fail to detect errors outside memory or in execution units, and current error detection methods introduce significant overhead when comparing results across machines.
Innovation Solution
A data processing device with signature generation circuitry that updates a signature independently of write order, allowing for compact error detection by performing bitwise operations between current signature values and data written to memory, enabling efficient comparison across machines without high overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If error detection codes are stored alongside message data in memory, then errors in memory cells can be detected, but errors outside memory (in execution units) will go undetected
Solution Approach 1:
The error detection function is segmented into two parts: error detection codes for memory data and a separate signature maintenance mechanism for execution unit outputs. The signature is updated independently based on write operations, allowing detection of errors in both memory and execution units without relying solely on memory-based error detection codes.
Solution Approach 2:
A signature register acts as an intermediary between the execution unit and the error detection system. Instead of directly monitoring execution unit outputs, the system maintains a signature that reflects the state of written data, providing indirect but effective error detection for both memory and execution unit errors.
2Reliability
If two copies of the application are run on separate machines and results are compared, then errors can be detected, but processing time increases significantly
Solution Approach 1:
The essential error detection information is extracted from the full application results into a compact signature. Instead of comparing entire application outputs between machines, only the compact signatures need to be compared, achieving error detection with minimal overhead and preserved productivity.
Solution Approach 2:
The error detection approach changes from comparing full result sets to comparing compact signature parameters. This parameter transformation reduces the data volume requiring comparison from potentially large application results to small signature values, maintaining reliability while improving productivity.
3Reliability
If full results from application execution are exported for comparison, then comprehensive error detection is achieved, but data transmission and processing overhead increases
Solution Approach 1:
The critical error detection information is extracted into a compact signature that can be exported and compared with minimal overhead. This signature captures the essential state of written data without requiring export of the entire application result set, significantly reducing time loss while maintaining error detection accuracy.
Data Source
AI summary
Signature generation circuitry is configured to update a signature in response to each of a plurality of writes to memory. The signature is updated by performing bitwise operations between current bit values of the signature and at least some of the bits written to memory in response a write. The bitwise operation are order-independent such that the resulting signature is the same irrespective of the order in which the writes are used to update the signature. The signatures are formed in an order-independent manner such that, if no errors have occurred in generating the data to be written to be memory, the signatures will match. In this way, a compact signature is developed that is suitable export from the data processing device for checking against a corresponding data processing device of a machine running a duplicate application.