Dallara uses HPC on AWS to offload maximum CFD workloads for race car simulations


In April 2021, Italian racing car manufacturer Dallara Automobili (Dallara) needed more high-performance computing (HPC) for simulation and testing than was available in its on-premises environment. The company’s computing power was overstretched, which made it difficult to meet the demands of its customers during peak season. As a major supplier of commercial race cars for prestigious championships, Dallara uses HPC to power the testing of its car designs, making HPC a fundamental part of its operations.

Dallara landed on Amazon Web Services (AWS) for the HPC he needed. Using AWS, Dallara built an HPC system that met its performance and cost benchmarks, leading the company to continue designing some of the fastest and most aerodynamic vehicles in the world. Dallara not only quickly found the solution to their critical problem on AWS, but also benefited from its scalability and flexibility.

Opportunity | Encounter a business-critical issue on-premises

Founded in 1972, Dallara manufactures race cars for the IndyCar, Indy Lights, Formula 2, Formula 3 and Super Formula championships. It produces cars for endurance racing like the 24 Hours of Le Mans and for electric car racing like Formula E. Today, Dallara is even developing road cars, attracting interest from luxury car makers . Each vehicle design is subjected to rigorous testing on the vehicle’s structure, aerodynamics and dynamics. For this, Dallara relies on more than 15 simulation and test tools that require massive amounts of HPC, including those that evaluate computational fluid dynamics (CFD). “We use CFD tools because it is mandatory to study the flow fields around our cars with all the necessary details to achieve our goal,” says Elisa Serioli, CFD Methodology Team Leader at Dallara.

Due to an influx of customer projects in February 2021, Dallara achieved 100% utilization of its on-premises HPC capacity. Serioli and the Dallara HPC team were tasked with upgrading the company’s HPC infrastructure and outsourcing its management to a cloud provider. “Our first goal was to have an industry-ready infrastructure that would support our specific applications, our huge models, and the high demand for HPC,” says Serioli. “The second goal was to integrate our workflows into an external environment like the cloud.”

Dallara sought proof of concept from various cloud providers, but AWS was the most responsive and supportive. Within a month of his request, Dallara was in production on AWS and running large-scale CFD simulations. “AWS support was there every day,” says Serioli. “AWS flexibility and commitment was key for us.” Additionally, Dallara was already using software from Ansys, an AWS partner, as its primary CFD solutions, in particular Ansys Fluent, a fluid simulation software. Another reason Dallara chose AWS is because they liked the ability to choose the right instance for each workflow using Amazon Elastic Compute Cloud (Amazon EC2), which provides secure and resizable compute capacity. for virtually any workload. For example, Dallara started using Amazon EC2 C5n instances, which are designed for compute-intensive workloads and use the fourth generation custom Nitro card and elastic network adapter to deliver 100 Gbps of network throughput to a single instance.

Solutions | Launch of a scalable HPC solution in less than 5 months

In April 2021, 2 months after the start of construction, Dallara had created an industrial infrastructure on AWS, united it with its existing workloads and allocated resources to it. The solution was stable and worked well within 5 months of heavy use. First, Dallara tied its on-premises workloads to AWS using Amazon Virtual Private Cloud (Amazon VPC), giving the company full control over its virtual network environment, including Amazon EC2 resource placement and AWS Virtual Private Network (AWS VPN) solutions that establish secure connections between on-premises networks, remote offices, client devices, and the AWS global network.

With its cloud and on-premises environments connected, Dallara decided to migrate 80% of its CFD workflow to the cloud and upload as little data as possible in order to delegate multiple tasks from each workflow to the cloud. “We use multiple software applications that each perform a different task for our complex CFD workflow, and the output of one job is the input of another,” Serioli explains. The connection between systems on AWS and on-premises facilitates a seamless user experience for Dallara aerodynamicists, who can choose where to perform each task or overall workflow. When a job runs in the cloud, the necessary files are automatically copied to Amazon FSx for Luster, which provides fully managed shared storage with the scalability and performance of the popular Luster file system. Then an orchestrator runs all the workflows. After each task is completed, the data is uploaded to the onsite solution and shared with aerodynamicists. Using FSx for Luster, Dallara can scale its file storage as needed in half an hour without any special assistance. On average, Dallara can run 15 full workflows per day.

Dallara leverages AWS ParallelCluster, an open-source cluster management tool that enables enterprises to easily deploy and manage HPC clusters on AWS. Using it, Dallara can immediately access additional HPC resources, scaling instances almost instantly…

Read the full case study to learn more. Reminder: You can learn a lot from AWS HPC engineers by subscribing to the HPC Tech Short YouTube channel and following the AWS HPC Blog channel.

Carol N. Valencia