Exosonic Uses Rescale on AWS and Intel to Rapidly Design Quieter, More Efficient Supersonic Aircraft
Executive Summary
Exosonic uses an HPC platform from AWS Partner Rescale to quickly scale and run complex aircraft simulations, eliminate application management and deployment time, and get its product to market faster. Exosonic relies on the Rescale platform, running on Amazon EC2 C5 instances powered by Intel processors, to run mission-critical engineering applications on AWS.
Seeking Better Performance and Scalability for Airflow Simulations
Supersonic aircraft, which can fly faster than the speed of sound, are not known for being quiet. Aerospace company Exosonic promises to change that. The company is currently building supersonic passenger planes that can fly with a muted sonic boom, as well as military aircraft that can enhance training capabilities. Exosonic is striving to design airliners that are faster than any currently available. “We want to build aircraft with the ability to fly anywhere in the world in half the time,” says Tim MacDonald, chief technology officer at Exosonic.
The Exosonic aircraft design process requires complex, computationally intensive engineering simulations, which the company runs on a high performance computing (HPC) platform hosted in a data center. For example, to simulate airflow around the aircraft, the company’s engineers run different scenarios to demonstrate how the aircraft will perform in specific situations. Because Exosonic wanted to bring its aircraft into production as quickly as possible, it sought a faster, more scalable solution for running simulations. “We didn’t have the financial or engineering resources to build our own HPC environment,” MacDonald says. “As a small startup, it was difficult for us to buy a bunch of computing clusters, and it wouldn’t be a good investment because we would only be fully utilizing the environment a fraction of the time.”
It wouldn’t be costeffective for us to build our own HPC environment and train people to manage it. In addition, it would be expensive to run a bunch of large simulations individually. We’re saving money by running on AWS with Intel technology.”
Tim MacDonald
Chief Technology Officer, Exosonic
Implementing the Rescale HPC Platform on AWS with Intel Technology
Exosonic realized the cloud would offer the scalability and elasticity it needed, and the company engaged Rescale, an AWS Partner, for assistance. Rescale provides a turnkey HPC platform on Amazon Web Services (AWS) that features a library of 1,000-plus commercial and open-source simulation applications for engineers and scientists. “We provide a turnkey HPC platform built for the cloud that makes it incredibly easy to deploy and run a large library of engineering and simulation applications on AWS,” says Kevin Cangemi, principal of strategic partnerships and alliances at Rescale.
The Rescale platform runs on AWS services including Amazon EC2 C5 Instances powered by Intel Xeon Scalable processors. “Intel produces highperformance processors that power simulation workflows for our customers” Cangemi says. The Rescale solution also relies on Amazon Simple Storage Service (Amazon S3) buckets to store engineering data.
Rescale’s FedRAMP-authorized platform also helps Exosonic comply with US International Traffic in Arms Regulations (ITAR), which control the manufacturing of defense-related products and services, by running on AWS GovCloud (US). “Our software has a number of restrictions with how it’s made and distributed, and we need to be careful about how we control the computational work we do,” MacDonald says. “Rescale helps us comply with regulations by optimizing configurations on AWS GovCloud.”
Credit: Exosonic
Scaling Easily to Run Complex Aircraft Simulations
Using Rescale on AWS backed with AWS Partner Intel technology, Exosonic can scale its HPC environment quickly and easily to support complex simulations that use more than 300 cores. “We couldn’t scale our aircraft simulations on our on-premises infrastructure as easily as we can with Rescale on AWS,” MacDonald says. For instance, the company’s engineers can simulate airflow coming toward an aircraft at different angles to determine how the plane will perform during takeoff and landing, or how it will perform over its mission overall. Engineers can also use Rescale to ensure that the airliner is quiet. “Normally, when an aircraft flies supersonically, there is a shock wave that comes off and it’s loud when it reaches the ground,” says MacDonald. “We’re shaping the aircraft so the shock waves don’t coalesce in quite the same way, which causes a quieter thump.”
Exosonic can access its compute resources on demand through Rescale’s simple webbased interface, without having to spend time setting up and managing a complex stack of software and hardware. “In just a few clicks, engineers using Rescale can accelerate the design process by running simulations on AWS without having to be experts on deploying and maintain HPC in the cloud,” says Cangemi. “Through Rescale, Exosonic has instant access to compute-optimized instances on AWS to run simulations at very large scale. Intel Xeon Scalable processors provide the performance and scaling potential that our customers value.”
Getting a Prototype into Production Faster
Exosonic is accelerating design and development by using Rescale to run as many simulations as possible. “In our onpremises HPC environment, we might have to run 500 simulations, with each one taking several hours,” MacDonald says. “If we only had access to one computer, we could only do one simulation at a time, which would take days to complete. By running the simulations on Rescale, we can finish them in a day.”
Using Rescale, Exosonic runs more simulations faster, which will ultimately help the company get its supersonic aircraft into production on schedule. “Running our simulations on Rescale, we can increase the number of iterative cycles we do, and we can complete them faster,” says MacDonald. “This will help us build our aircraft faster.”
Optimizing Operational Costs
By avoiding the need to build its own internal HPC platform and taking advantage of the elasticity of AWS, Exosonic can more easily control its operational costs. “It wouldn’t be costeffective for us to build our own HPC environment and train people to manage it,” MacDonald says. “In addition, it would be expensive to run a bunch of large simulations individually. We’re saving money by running on AWS with Intel technology.”
Exosonic plans to continue using Rescale to run simulations for all of its projects, including the current focus on smaller scale aircraft for military training. MacDonald concludes, “Rescale and AWS give us the scalability and compute power we need, which will help us keep accelerating innovation.”
About Exosonic
Exosonic, based in Los Angeles, California, is an aerospace company that is developing supersonic aircraft, such as supersonic uncrewed aerial vehicles, to address commercial and government missions like air combat training. The company is designing the world’s first quiet supersonic airliner, which will be capable of traveling anywhere around the world in half the time.
AWS Services Used
Benefits
- Quickly scales and runs complex aircraft simulations
- Eliminates application management and deployment time
- Gets product to market faster
About the AWS Partners
Rescale, a technology company and AWS Partner, offers high performance computing built for the cloud to empower engineers while giving IT security and control. From supersonic jets to personalized medicine, industry leaders are bringing new product innovations to market with speed and efficiency with Rescale, a cloud platform delivering intelligent full stack automation and performance optimization.
Intel is the world’s leading designer and manufacturer of high-performance processors for servers, PCs, IoT devices, and mobile devices. AWS and Intel engineers have worked together for more than 10 years, building custom hardware to ensure AWS services run on a platform optimized for customer workloads for the best value. Intel Xeon Scalable processors power Amazon EC2 instances to help enterprises drive performance for their compute-intensive workloads.
Published April 2023