Instruction Execution Cost Mapping for Fast, Explainable Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining execution cost of instructions are either expensive and time-consuming when executed on actual hardware, inaccurate when using models, or computationally intensive and lack explainability with AI-based techniques.
Innovation Solution
Developing architecture-specific mapping tables that map converted instructions to average execution cost and standard deviation, allowing for efficient estimation of execution cost on different computer architectures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If execution cost is determined by executing instructions on actual hardware, then measurement precision is improved, but loss of time and use of energy increase
Solution Approach 1:
The patent pre-calculates and stores execution costs for different instruction types in a lookup table during system initialization or offline processing. When execution cost determination is needed, the system simply queries the pre-computed table rather than executing instructions on hardware, achieving both accurate measurement and fast retrieval.
Solution Approach 2:
The patent creates a simplified model or representation of execution cost data that can be queried without actual hardware execution. The lookup table serves as a copy or approximation of the complex hardware execution behavior, allowing fast queries that mimic actual measurement results.
2Measurement precision
If execution cost is determined by executing instructions on actual hardware, then measurement precision is improved, but use of energy increases
Solution Approach 1:
The system performs energy-intensive execution cost measurements during offline processing or system initialization, storing results in a lookup table. During actual operation, the system queries this table with minimal energy expenditure, avoiding repeated hardware execution and associated energy consumption.
3Productivity
If AI-based techniques are used to estimate execution cost, then productivity is improved, but device complexity increases
Solution Approach 1:
The patent uses simple, lightweight data structures (lookup tables with instruction type and execution cost pairs) instead of complex AI models. This approach provides sufficient accuracy for execution cost estimation while maintaining system simplicity and avoiding the computational overhead of AI-based techniques.
Data Source
AI summary
Methods, apparatus, systems, and articles of manufacture to determine execution cost are disclosed. An example apparatus includes memory; instructions included in the apparatus; and processor circuitry to execute the instruction to: cause a plurality of instructions corresponding to a mnemonic to be executed; determine an average execution cost of the plurality of instructions; determine a standard deviation of execution costs of the plurality of instructions; and generate a mapping table including an entry, the entry including the mnemonic in association with the average and the standard deviation.