Processor Fault Reproduction Time Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods fail to accurately estimate the time required for a CPU to reach a fault occurrence temperature, making it difficult to report the execution time of a reproduction test to a job scheduler in parallel computing systems, due to variations in CPU type and thermodynamic properties.
Innovation Solution
A processor-fault reproduction method that determines the heating time by using a heating program to heat the CPU to a fault occurrence temperature based on specific information about the CPU's current temperature, power consumption, refrigerant temperature, and fault occurrence temperature, and then adds the fault reproduction time to calculate the total execution time, which is reported to the job scheduler.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a heating program is executed to heat the processor to fault occurrence temperature, then the fault reproduction accuracy is improved, but the execution time estimation becomes inaccurate due to variations in CPU type and thermodynamic properties
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing heating time information for multiple CPU types and power consumption combinations before actual fault reproduction. The system stores heating time from current temperature to fault occurrence temperature in advance, so when a fault occurs, the pre-computed heating time can be directly retrieved and combined with fault reproduction time to provide accurate total execution time estimation without needing to recalculate during actual fault reproduction.
Solution Approach 2:
The patent utilizes parameter changes by considering multiple CPU types with different thermodynamic properties and power consumption values. The system pre-calculates heating times for various combinations of CPU types, power consumption levels, and temperature ranges, storing these as lookup data. When a fault occurs, the appropriate pre-computed heating time is selected based on the actual CPU type and power consumption, ensuring accurate execution time estimation despite variations in thermodynamic parameters.
2Measurement precision
If the heating time is calculated based on multiple parameters (power consumption, temperature, refrigerant temperature), then the execution time estimation accuracy is improved, but the complexity of the calculation system increases
Solution Approach 1:
The patent reduces calculation system complexity by applying preliminary action - all complex heating time calculations based on multiple parameters (CPU type, power consumption, temperature differential, refrigerant temperature) are performed in advance and stored in a database. The runtime system only needs to retrieve pre-computed heating time values based on matching parameters, avoiding complex real-time calculations while maintaining high accuracy.
Solution Approach 2:
The patent uses copying by creating a comprehensive lookup table that copies and stores heating time data for various CPU types and operating conditions. Instead of performing complex thermodynamic calculations during fault reproduction, the system copies pre-computed heating time values from the stored data structure, matching the actual CPU type and power consumption to retrieve the appropriate heating time estimate.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This method allows for accurate estimation of the heating time and reporting of the execution time for a reproduction test, enabling efficient scheduling and fault reproduction in parallel computing systems.
Implementation Method 1
a heating program which heats the processor to a fault occurrence temperature when a fault occurs from a current temperature
Implementation Method 2
a refrigerant temperature of a cooling medium to cool the processor
Data Source
AI summary
A processor-fault reproduction method includes: determining a heating time of a processor taken using a heating program which heats the processor to a fault occurrence temperature when a fault occurs from a current temperature based on first information regarding the current temperature of the processor, second information regarding a power consumption value and a temperature of the processor before the fault occurs and a refrigerant temperature of a cooling medium to cool the processor, third information regarding the fault occurrence temperature, and fourth information regarding a power consumption value of the processor during execution of the heating program; determining an execution time by adding a fault reproduction time taken by a fault reproducing program which reproduces a state of the processor when the fault occurs to the heating time; and reporting the execution time to a job scheduler with a request to execute the heating and fault reproducing programs.