Enhance EDA engineering team’s experience and lower costs
EDA Workload Management
For EDA Engineering teams, misconfigured queueing systems or a lack of intelligent queueing systems can cause inconsistent user experiences. These issues lead to underutilized cluster resources and unknown job execution times, decreasing productivity and experience. Over provisioned EDA Software Licenses are the most critical aspect of EDA optimization which provides tangible value to customers. Axomo for EDA Workloads identifies and fixes misconfigurations, implements intelligent queueing, and optimizes EDA Licensing requirements to reduce operational costs, enhance the user experience and provides predictability of cluster resources.
Comprehensive support for EDA resource utilization
Axomo for EDA Workloads on AWS provides descriptive analytics on: Job wait time, execution time, queue consumption, recommended queue structure, and ways to optimize wait and execution times. Use this AWS Quick Start to deploy an Amazon Cloud Formation Template in an existing VPC and connect to on-premise IBM LSF, RTM, and Telegraf servers. Then transform and load training data (a minimum of 24 months) on an Amazon Simple Storage Service (Amazon S3) backed data lake. AWS Database Migration Service (AWS DMS) uses the change data capture mechanism to incrementally update the data lake on an hourly or daily basis. AWS Cloud Formation also deploys Amazon Glue to perform data preparation tasks for an Amazon Sagemaker model to enable intelligent queuing, recommended fixes, and predict resource requirements for future projects.
“We were able to optimize and confidently re-architect the HPC Cluster for better resource utilization, enhanced job performance, and improve the throughput of one of our largest EDA cluster by up to 40%, and improve queue contention related to jobs having lesser ‘wait/pending’ times using smart queues. The AI-driven forecasting model will help us predict future resource requirements and accelerate our cloud journey.”
– Michael Brooker, CIO at Synaptics
Fix misconfiguration, underutilization, and over provisioning of EDA resources
Intuitive Dashboard View
Axomo for EDA Workloads includes a pre-built dashboard view that provide insights for job types and underutilized hardware and misconfiguration.
Reduce EDA License Costs
Axomo for EDA Workloads key outcome is to support the reduction of over provisioned EDA Software Licenses and provide future licensing guidance.
Automate and Simplify
Axomo for EDA Workloads supports AI enabled queueing for job assignment to optimize cluster and license utilization.
Recommended Actions or “Fixes”
Axomo for EDA Workloads helps customers right size AWS Batch and Amazon EC2 Spot Fleet resources which help customers reduce EDA cluster infrastructure costs by 40% and provide better cost predictability.
Axomo for EDA Workloads Supports:
- A structured data lake in Amazon S3 to hold the raw, modeled, enhanced, and transformed data.
- A staging bucket for the feature engineered and transformed data that will be ingested into Amazon SageMaker.
- Data transformation code hosted on AWS Lambda to prepare the raw data for consumption and ML model training, and to transform data input and output.
- Amazon SageMaker automation through AWS Lambda functions to build, manage, and create REST endpoints for new models, based on a schedule or triggered by data changes in the data lake.
- An AWS SageMaker notebook server to enable data exploration using a Jupyter notebook.
- AWS Identity and Access Management (IAM) to enforce the principle of least privilege on each processing component. The IAM role and policy restrict access to only the resources that are necessary.
- A demo scenario that builds and updates a predictive model for Amazon QuickSight consumption
- AWS Data Migration Service reads from on-premise IBM RTM® MySQL Database to migrate the data to Amazon S3 over Secured VPN
- Amazon Simple Workflow Service (Amazon SWF) reads IBM LSF® and Telegraf Logfiles to transfer them over Secured VPN to Amazon S3
Case Study: Synaptics
Synaptics, a leading human machine interface manufacturer in Silicon Valley, used IBM LSF, RTM, and Telegraf to operate and manage EDA workloads for their internal design and testing teams. These teams needed to optimize their EDA resource utilization and enhance their EDA team’s resource management experience.
Synaptics’ EDA Team had challenges managing resource utilization and execution timelines of LSF cluster for jobs executed by multiple teams across 6 global locations.
The Synaptics EDA team used Axomo for EDA Workloads dashboards powered by machine learning data models to predict job runtime and resource utilization.
Synaptics EDA team achieved 70%+ improved efficiency of LSF cluster. Business units are now able to execute projects 10x -12x faster with smart queues.