Cloud EDA Job Scheduling Using ML Resource Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for managing electronic design automation (EDA) on cloud face challenges such as identifying the best cloud service provider, optimizing resource utilization, determining suitable jobs for cloud bursting, and predicting costs, leading to inefficiencies like costly burst patterns and server farm tilt due to static priority-based policies.

Innovation Solution

A method and system using machine learning models to predict optimal resource configurations and configuration circuits on cloud, evaluating different cloud service providers for lowest cost, calculating job completion and burst times, and determining wait times to deploy EDA jobs efficiently on-premises or on cloud, thereby overcoming inefficiencies.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If cloud bursting is used to handle volume of design simulation jobs, then job throughput is improved, but cost increases and resource utilization becomes inefficient

Engineering Contradiction:

Improvejob throughputVSAvoidcost

Core Design Contradiction:

ProductivityVSLoss of energy

Solution Approach 1:

The system dynamically adjusts resource allocation between on-premises server farm and cloud infrastructure based on real-time job queue conditions, utilization metrics, and cost parameters. The hybrid architecture enables flexible scaling where jobs are processed on-premises when resources are available and burst to cloud when needed, optimizing both throughput and cost efficiency

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes operational parameters by monitoring utilization thresholds, cost parameters, and job characteristics to determine optimal deployment locations. By adjusting these parameters dynamically, the system achieves cost-effective resource utilization while maintaining high job throughput through intelligent cloud bursting decisions

Inventive Principle:

Principle #35Parameter changes

2Ease of operation

If static priority-based policies are used for job scheduling, then implementation is simple, but resource utilization becomes inefficient and server farm tilt occurs

Engineering Contradiction:

Improvescheduling simplicityVSAvoidresource utilization efficiency

Core Design Contradiction:

Ease of operationVSProductivity

Solution Approach 1:

The scheduling system transitions from static priority-based policies to dynamic multi-parameter decision-making that considers job characteristics, resource availability, cost parameters, and utilization metrics in real-time, optimizing resource allocation without sacrificing operational simplicity

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system implements feedback mechanisms by continuously monitoring job completion rates, resource utilization, and cost parameters, then using this information to adjust scheduling decisions and prevent server farm tilt, improving overall resource utilization efficiency

Inventive Principle:

Principle #23Feedback

3Productivity

If cloud resources are used to meet design simulation demands, then processing capacity is improved, but identifying best fit cloud service provider becomes challenging

Engineering Contradiction:

Improveprocessing capacityVSAvoidcloud service provider selection complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system creates a universal interface layer that abstracts the complexity of multiple cloud service providers, enabling the hybrid architecture to work with different CSPs through standardized protocols and metrics, simplifying provider selection while maintaining access to diverse processing capacities

Inventive Principle:

Principle #6Universality (Multi-functionality)

4Productivity

If server farm size is increased to handle job volume, then processing capacity is improved, but adaptability to varying job requirements decreases

Engineering Contradiction:

Improveprocessing capacityVSAvoidadaptability to job requirements

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The system segments processing capacity into on-premises server farm and cloud infrastructure components, allowing independent scaling and adaptation of each segment to match specific job requirements, thereby maintaining both high processing capacity and flexibility

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS12517758B2Method and system for managing electronic design automation on cloud

Publication Date: 2026.01.06 TATA CONSULTANCY SERVICES LTD

AI summary

Existing techniques of managing Electronic Design Automation (EDA) on cloud are based on pre-defined policies which result in costly burst patterns and server farm tilt. Embodiments of present disclosure overcomes these drawbacks by a method and system for managing EDA on cloud which employ machine learning to predict optimal resource configurations for deploying EDA jobs and configuration circuit on cloud that holds resources required by the optimal resource configuration. Further, different Cloud Service Providers (CSPs) are evaluated to determine the least cost CSP which has the desired configuration circuit. Completion time of jobs, time required to burst the jobs on cloud, and a pre-defined desired cycle time are calculated as a sum to determine the corresponding wait time for each job. The jobs are retained in the queue for corresponding wait time before deploying them on the cloud. The jobs are deployed on the on-prem infrastructure if resources are freed up before the wait time.