Normalized Performance Units for Cross-Platform Computing Benchmarking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current benchmarking techniques fail to provide a standardized, normalized measurement system for evaluating and improving processing performance across multiple computing resources and platforms, especially when dealing with disparate technologies and fluctuating costs, making it difficult for CTOs and CIOs to make informed decisions about infrastructure management.
Innovation Solution
A computer-implemented method and system that calculates a normalized benchmark for each computing resource, determines performance units, and provides a graphical user interface for visual comparison, allowing for the evaluation of processing performance indicators and cost analysis across multiple computing resources, enabling informed decision-making.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional benchmarking techniques are used to assess performance, then relative performance can be measured for individual systems, but a standardized measurement system across multiple disparate platforms cannot be established
Solution Approach 1:
The patent transforms raw benchmark scores into normalized performance units by applying mathematical transformations. The system calculates performance units using the formula: Performance Units = (Benchmark Score / Reference Score) × Reference Performance Units, where the reference represents a standardized baseline system. This parameter transformation enables consistent measurement across disparate platforms while maintaining adaptability to different benchmarking tools and methodologies.
Solution Approach 2:
The patent introduces performance units as an intermediary metric that mediates between different benchmarking standards and platforms. Instead of directly comparing raw scores from different benchmarks (which measure different things in different units), the system converts all measurements through the performance unit intermediary, enabling standardized comparison while preserving the characteristics of each platform.
2Measurement precision
If manual calculations are performed to determine performance values, then detailed analysis can be conducted, but the process requires many man-hours and is inefficient
Solution Approach 1:
The patent implements automated calculation engines that perform performance unit determinations without manual intervention. The system automatically collects benchmark data, applies normalization algorithms, calculates performance units, and generates comparative analyses. This self-service automation maintains the precision of detailed analysis while eliminating the manual labor requirements, reducing decision-making time from many man-hours to automated processing.
Solution Approach 2:
The patent replaces manual mechanical calculation processes with automated computational systems. Instead of personnel manually collecting data, performing calculations, and generating reports, the system uses software-based calculation engines that automatically process benchmark data through normalization formulas and generate performance comparisons, maintaining analytical precision while dramatically reducing time investment.
3Productivity
If benchmark scores are used directly for comparison, then raw performance data is available, but normalization across different architectures and platforms cannot be achieved
Solution Approach 1:
The patent creates an equipotential measurement field by normalizing all performance measurements to a common reference level. The system establishes a reference system with known performance characteristics and adjusts all other measurements relative to this reference, ensuring that comparisons are made under equivalent conditions. This equipotential approach enables efficient productivity comparisons while maintaining measurement precision across different architectures by eliminating platform-specific biases.
Data Source
AI summary
The invention is directed to a computer-implemented method and system for improving processing performance for a group of computing resources, the method implemented on at least one computer having a processor and accessing at least one data storage area. The method comprises implementing the processor for calculating a benchmark for each computing resource in the group of computing resources and normalizing the benchmark across the group of computing resources to determine a number of performance units for each computing resource. The method additionally includes providing a graphical user interface for facilitating visual comparison for comparing processing performance indicators for multiple computing resources in the group of computing resources and reconfiguring at least some of the computing resources represented on the graphical user interface based on the comparison.