Workload Model Generation Using LSTM Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Information handling systems face challenges in identifying and managing diverse and dynamic workloads across different user systems and industries, making it difficult to ensure they meet user needs effectively.
Innovation Solution
The method involves leveraging queuing theory and Long Short-Term Memory (LSTM) networks combined with Machine Learning (ML) to analyze workload data from hundreds or thousands of user systems, dividing it into bins based on characteristics, clustering, and generating a workload model that can be deployed to ensure information handling systems meet user requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If workload data from hundreds or thousands of user systems is analyzed manually or using traditional methods, then the analysis process becomes complex and time-consuming, but the ability to identify emerging workloads and generate accurate workload models is improved
Solution Approach 1:
The patent replaces manual analysis and traditional processing methods with machine learning models (LSTM networks) and automated clustering algorithms. This substitution enables the system to automatically identify emerging workloads and generate workload models from large volumes of data without requiring manual intervention, thereby maintaining high identification accuracy while reducing processing complexity.
Solution Approach 2:
The patent creates workload models that replicate actual workload patterns observed from hundreds or thousands of user systems. These models serve as simplified representations that capture essential workload characteristics without requiring direct analysis of the original complex data, enabling efficient validation and demonstration of information handling system capabilities.
2Adaptability or versatility
If traditional workload analysis methods are used, then the process is simpler to implement, but the system cannot effectively identify emerging workloads across diverse user systems and industries
Solution Approach 1:
The patent develops workload models with a universal structure that can accommodate diverse workload types across different user systems and industries. The model framework is designed to be industry-agnostic and workload-type-agnostic, allowing it to adapt to various scenarios including storage, networking, and computing workloads from different sectors such as finance, healthcare, and enterprise environments.
Solution Approach 2:
The patent employs automated machine learning pipelines that automatically process workload data, perform clustering analysis, and generate workload models without manual configuration. This automation enables the system to handle the complexity of diverse workloads across multiple industries while maintaining consistency and scalability in model generation.
3Manufacturing precision
If workload data is not segmented and clustered, then the data processing is simpler, but the ability to determine accurate workload mixes for different information handling systems is reduced
Solution Approach 1:
The patent segments workload data into distinct bins based on multiple characteristics such as workload type, industry, system size, and performance metrics. This segmentation allows the system to analyze and model different workload categories separately, leading to more accurate workload mix determinations for specific information handling systems while managing complexity through structured data organization.
Solution Approach 2:
The patent applies different clustering approaches and analysis methods tailored to specific workload categories and industry segments. Rather than applying a uniform analysis method to all data, the system adapts its processing approach to local characteristics of different workload types, improving the precision of workload mix determination for each specific context.
Data Source
AI summary
A system for generating a workload model for a target information handling system includes gathering workload data from a plurality of information handling systems, dividing the workload data into a plurality of workload data bins, identifying workload data characteristics for each workload data bin and identifying workload data sets that may be applicable to the target information handling system. A workload mix may be determined based on the target information handling systems and workload data characteristics. Real customer workload data including real-time or near real-time workload data may be used to check a workload model for accuracy before deploying the workload model to a target information handling system.