BeyondMath is changing Design with AI-driven CFD Simulations that could drastically reduce wait times and accelerate the design process.
“Unlike language, where we don’t have mathematical models to describe what the next word should be, when it comes to physics, we do have those models,” explains BeyondMath’s co-founder Darren Garvey. “And what we’re seeing is that machine learning is actually quite good at computation, not just pattern recognition.”
BeyondMath’s initial focus is on computational fluid dynamics (CFD), a field essential for predicting how objects interact with air or water. While CFD has been a staple in industries like automotive and aerospace, it has traditionally required immense computational power and time, often involving supercomputers and GPU clusters. Designers typically have to wait overnight to see the results of their simulations, slowing down the iteration process significantly.
BeyondMath’s goal is to speed up this digital design phase by shortening the feedback loop between ideation and validation. Garvey illustrates the potential impact: “Imagine you’ve got six months to design a part for a plane. Given that a simulation takes so long, you might get 20 attempts to try things out. But if a designer thinks of an idea and gets results within seconds or a couple of minutes, in that same six months you might be able to run a million changes.”
Their first product, a “digital wind tunnel”, offers near-real-time simulation of airflow over complex surfaces. This would typically take hundreds of times longer using conventional methods. The approach draws parallels with machine learning applications in weather prediction, where models can simulate systems quickly and accurately without relying solely on pre-existing datasets.
BeyondMath doesn’t rely on traditional simulations alone; instead, the company’s model integrates a deep understanding of both theoretical principles and real-world data. “We’re not trying to approximate the simulations, we’re trying to approximate the real world,” Garvey emphasizes. This unique approach allows BeyondMath’s model to actively participate in the design process, offering insights that traditional CFD methods cannot match.
Formula 1 racing, known for its cutting-edge technology and rapid adoption of new innovations, is among BeyondMath’s first markets. Several F1 teams are already evaluating the software to enhance their aerodynamics and vehicle design processes. “They’re one of the heaviest users of CFD, and they’re fast-moving, they’ll adopt new technologies. We’re close to having a platform that will actually make their cars faster,” Garvey notes.
BeyondMath recently raised $8.5 million in a seed round led by UP.Partners, with Insight Partners and InMotion Ventures also participating. The funding will help the startup double its team and scale up its computing power, including purchasing Nvidia DGX 200s, as they collaborate with Nvidia on this pioneering application of their hardware.
Looking ahead, BeyondMath aims to expand beyond its initial focus on top-tier customers like Formula 1 teams. However, Garvey acknowledges that generalizing their models to different domains will be challenging. “For example, if a model understands cars, or car-like objects, it’s not necessarily going to understand a plane, or a blood vessel,” he says. But with a clear path to traction, BeyondMath is poised to revolutionize how industries approach design and simulation.
read more: ElevenLabs Expands AI-Powered Reader App Globally with Support for 32 Languages
