High-Performance Computing Predictive Diagnostics via AI Emulation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Diagnostic tests in high-performance computing environments are time- and resource-intensive when executed at high utilization levels, leading to potential hardware wear, undetected issues, unproductive resource use, and a significant carbon footprint, while infrequent testing may result in undetected problems and compromised performance.

Innovation Solution

Perform diagnostic tests at lower restriction levels, generating empirical parameter values, and use AI models to infer performance at higher levels, reducing physical stress and resource consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If diagnostic tests are executed at high utilization levels, then performance evaluation accuracy is improved, but resource consumption and hardware wear increase

Engineering Contradiction:

Improveperformance evaluation accuracyVSAvoidresource consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent creates a virtual diagnostic environment that copies and emulates diagnostic test behavior through AI models. Instead of physically executing resource-intensive diagnostic tests on hardware, the system uses machine learning models trained on diagnostic data to simulate and predict diagnostic outcomes, thereby obtaining accurate performance evaluations without the actual hardware stress and energy consumption of running full diagnostic tests.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The system changes the fundamental parameters of how diagnostics are executed by transitioning from physical test execution to virtual prediction. The AI models process input parameters (hardware specifications, operational conditions) and output predicted diagnostic results, replacing the traditional approach of physically measuring performance metrics through resource-intensive diagnostic routines.

Inventive Principle:

Principle #35Parameter changes

2Reliability

If diagnostic tests are executed frequently, then detection reliability is improved, but hardware stress and resource consumption increase

Engineering Contradiction:

Improvedetection reliabilityVSAvoidhardware stress

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The patent employs AI models that copy the diagnostic evaluation function without requiring physical execution. The models are trained on diagnostic data and can repeatedly predict performance outcomes without imposing hardware stress. This enables frequent diagnostic evaluations to be performed virtually, maintaining high detection reliability while eliminating the harmful effects of repeated physical testing on hardware.

Inventive Principle:

Principle #26Copying

3Measurement precision

If diagnostic tests are executed at high utilization levels, then performance measurement accuracy is improved, but carbon footprint increases

Engineering Contradiction:

Improveperformance measurement accuracyVSAvoidcarbon footprint

Core Design Contradiction:

Measurement precisionVSLoss of energy

Solution Approach 1:

The system replaces energy-intensive physical diagnostic tests with virtual AI-based prediction. The machine learning models process data and generate performance predictions without requiring the hardware resources that would otherwise be consumed during actual diagnostic test execution, thereby maintaining measurement accuracy while dramatically reducing energy consumption and carbon footprint.

Inventive Principle:

Principle #26Copying

4Use of energy by moving object

If diagnostic tests are executed at lower restriction levels, then resource consumption is reduced, but measurement precision decreases

Engineering Contradiction:

Improveresource consumptionVSAvoidperformance evaluation accuracy

Core Design Contradiction:

Use of energy by moving objectVSMeasurement precision

Solution Approach 1:

The system fundamentally changes the measurement approach by using AI models that can accurately predict diagnostic outcomes without requiring high-utilization test execution. The models learn the relationships between hardware parameters, operational conditions, and diagnostic results during training, enabling them to provide precise performance evaluations based on input data rather than requiring resource-intensive physical measurement.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250328442A1Predictive diagnostics in high-performance computing

Publication Date: 2025.10.23 HEWLETT PACKARD ENTERPRISE DEV LP

AI summary

A development system for predictive diagnostics is provided. During operation, the system can perform a first diagnostic test on a distributed computing system based on a first restriction level indicating resource consumption of a first set of hardware units. The distributed computing system can include a plurality of computing devices with processing and memory resources. The system can generate a first log comprising a first set of parameter values indicating an output of the first diagnostic test at the first restriction level of the distributed computing system. The system can configure a first diagnostic tool with the first set of parameter values to emulate the first diagnostic test. The system can then apply the first diagnostic tool to obtain a second set of parameter values indicating an output of the first diagnostic test at a second restriction level, which can be higher than the first restriction level.