proteanTecs Supports Tens of Millions of Simulations, Cuts Costs by Up to 60% Using AWS

A Startup Seeks Flexibility and Speed

proteanTecs is a startup company and provider of a software solution that uses deep data and machine learning to predict failures in electronics. Microchip manufacturers use the solution to combine data, derived from agents (IPs) embedded in chips, with predictive artificial intelligence (AI) and data analytics to track faults before they become failures, thus increasing quality and reliability of mission-critical electronics. The company also develops electronic design automation (EDA) software tools to support its solution.

As part of its development efforts, proteanTecs needed the flexibility and scalability to support high-performance computing (HPC) workloads that run millions of simulations each day. “We process a large volume of data for each chip and then process machine-learning algorithms from that data,” says Yuval Bonen, co-founder and vice president, software for proteanTecs. “We might need to run tens of millions of simulations one week and only a few million the following week,” Bonen says. “We really need the flexibility to accommodate our changing requirements.” The company must also meet its customers’ deadlines for EDA software delivery. “We have to deliver projects on time every time,” says Bonen. As a startup, proteanTecs also wanted to reduce its operational costs as much as possible while remaining prepared for rapid business growth.

 

“Based on our customers’ requirements, we may need to run tens of millions of simulations in one week and only several million the next week. AWS gives us the flexibility to do that without any effort at all.”

-  Yuval Bonen, Co-founder and Vice President, Software, proteanTecs

  • About proteanTecs
  • Benefits
  • AWS Services Used
  • About proteanTecs
  • proteanTecs develops revolutionary Universal Chip Telemetry for electronic systems throughout their entire lifecycle, increasing their performance and reliability. By applying machine learning to novel data created by embedded agents, proteanTecs provides meaningful insights unattainable until today, leading to new levels of quality, reliability, and scale. Founded in 2017 and led by industry veterans, the company is headquartered in Israel with offices in New Jersey and California.

  • Benefits
    • Runs tens of millions of simulations on Amazon EC2 C5 instances
    • Uses Amazon EC2 Spot Instances to save up to 60% annually
    • Enables focus on researching and developing new products
  • AWS Services Used

Running SaaS and EDA Solutions on AWS

proteanTecs conducted a proof of concept (POC), with assistance from Amazon Web Services (AWS) architects, which showed how a migration to the AWS Cloud would increase performance and scalability. “AWS worked closely with us throughout the POC to help us harness the benefits of the cloud in EDA workloads,” says Bonen. “We realized that AWS technology was a great fit for us in terms of performance, reliability, and flexibility.”

Following the POC, proteanTecs moved its HPC workloads to AWS, running on Intel Xeon processor–powered Amazon Elastic Compute Cloud (Amazon EC2) C5 Instances. The company uses Amazon Relational Database Service (Amazon RDS) to store application data and runs a Kubernetes container orchestration system on Amazon EC2.

To optimize costs, proteanTecs recently began using Amazon EC2 Spot Instances to support its workloads. Spot Instances are spare Amazon EC2 capacity offered at up to a 90 percent discount compared to On-Demand prices.

Performing Tens of Millions of Simulations in Parallel Weekly

By using AWS, proteanTecs can scale to run tens of millions of simulations in parallel every week. “We can easily support our heavy compute and memory requirements by running our HPC workloads on Amazon EC2,” says Bonen. The company also has the flexibility to automatically scale up or down. “Based on our customers’ requirements, we may need to run tens of millions of simulations in one week and only several million the next week,” says Bonen. “AWS gives us the flexibility to do that without any effort at all.”

Speeding Time-to-Market

proteanTecs can quickly provision new compute and storage resources, which speeds time-to-market for delivering its software-as-a-service (SaaS) and EDA solutions to customers. “Running on AWS, we can provision resources within minutes, which helps us accelerate time-to-delivery for our customers,” says Bonen. “For example, we had a customer that was pushing us to deliver a project two weeks earlier than scheduled, and we met that deadline—something we could not have done without AWS. We can scale quickly and consume twice as many compute and storage resources on AWS than we could in an on-premises environment.”

Cutting Costs by 60%

By taking advantage of Amazon EC2 Spot Instances, proteanTecs is running its environment more cost-effectively. “We are reducing our operational costs by up to 60 percent using Amazon EC2 Spot Instances,” says Bonen. “We use those savings to invest more resources into research and development and create new product offerings that will differentiate our company.”


Learn More

Learn more about high performance computing on AWS.