Some of the greatest ideas were born in humble garages. Airplanes, rock-and-roll, and that longest of longshots, the personal computer, all ignited full-on revolutions. So, Mate Rimac felt in good company when he set out to design and build the world’s most powerful and advanced electric vehicle in his own garage in Croatia. He knew plenty of doubters—people who told him it would be impossible to create a brand-new car company in a country with such a small automotive industry—but Rimac wanted to revolutionize the industry from the ground up, so he knew it didn’t matter where he started.
However, complex the model we need to create, we know that we can manage it with Azure HPC. We now produce more highly complex models that simply wouldn’t have been possible on our old infrastructure.
Ivan Krajinović: Head of Simulations, Rimac Technology
From the garage to greatness
And he’s proved the doubters wrong. Rimac Technology builds and sells the Nevera—named for the sudden, furious, lightening-charged storms that break out over Croatian seas. This near 2,000-horsepower electric sports car can be fully charged in as little as 20 minutes, and it’s the world’s fastest accelerating production car, reaching the record time of 8.582 seconds per quarter mile at a top speed of 167.51 mph. Rimac also supplies leading-edge electric vehicle technology and components to other manufacturers, including Porsche and Bugatti.
It didn’t come easy. Making it all happen has taken plenty of brilliance, commitment, hard work, and digital horsepower. Rimac recently began using cloud computing to help with the design, testing, and manufacturing of its next-generation components and sportscars, and it’s gaining even greater scale and speed in its product development processes with a boost from Microsoft Azure HPC underpinned by AMD EPYC™ CPUs.
Success through simulation
A central plank of Rimac’s ability to design and realize such high levels of performance in its components and vehicles is its simulations. The company performs simulations for core components like electric motors, inverters, gearboxes, and the battery packs that power both its own cars and other manufacturers’ cars. These simulations enable a level of focused and precise data collection that can’t be matched through physical testing, where something as simple as a misplaced sensor runs the risk of compromising data, or worse yet, damaging a €2 million car.
Ivan Krajinović, Head of Simulations at Rimac Technology, states, “The power of simulation is that you can build really complex models, test them virtually, and then iterate and modify the design to do the simulation again, all without needing to produce anything.” Simulations yield incredibly granular data that Rimac couldn’t access or process through physical testing. Krajinović continues, “When a car crashes against a wall, you can’t see, for example, that a bracket somewhere in the middle of the car is deformed and the source of a larger problem. That’s only identifiable through the power of digital virtual prototyping.” The company needs this level of data collection and complex processing to build powerful electric vehicles like the Nevera.
High-performance engineering meets high-performance computing
Initially, Rimac developed its own internal HPC cluster to provide the computing resources and storage it needed to innovate and iterate its products. However, as Rimac grew, it required greater scale and more powerful computing. Migrating workloads from its internal cluster to the cloud offered the prospect of improving simulations and models, adding greater complexity to those models, and doing it all faster.
Krajinović recounts, “We needed more computational power to work with more complex simulation models to match increasingly challenging customer requests. Our internal cluster was really strong, but the availability of the nodes and the calculations we could run with Azure HPC were just amazing.”
Also, supply chain challenges would have made the addition of physical nodes to Rimac’s internal cluster far more time-consuming than a cloud solution. “The estimate to get additional physical nodes was 9 to 12 months, and we would have needed 3 more months to install them,” says Krajinović. “Adding a node in an Azure HPC environment, on the other hand, is basically instant.”
On top of the powerful infrastructure and resources harnessed from Azure, Rimac is running its engineering simulations on UberCloud, a cloud- and application-agnostic platform that acts as an interface between cloud high-performance computing (HPC) infrastructure and engineering workflows. A member of the Microsoft Partner Network, UberCloud runs simulation workflows in closed, highly secure software containers. So, Rimac engineers run whichever application container or simulation workflow they need, and then it gets loaded onto the Azure HPC cluster.
Rimac’s Azure HPC environment uses Azure CycleCloud to organize and orchestrate clusters—putting together different cluster types and sizes flexibly and as necessary. The solution includes Azure Virtual Machines, running containers on Azure HBv3 virtual machines with 3rd Gen AMD EPYC™ Milan Processors with AMD 3D V-Cache, which are much faster than previous generation Azure virtual machines for explicit calculations. Rimac’s solution takes full advantage of the power of AMD, which offers the highest performing x86 CPU for technical computing. “We’ve gained a significant increase in computational speed with AMD, which leads to lower utilization of HPC licenses and faster iterations,” says Krajinović.
Better, faster, more competitive
By shifting workloads to Azure HPC, Rimac enjoys greater scale and flexibility with fewer headaches. Rimac need not worry about resources or having to acquire and set up new hardware. Krajinović states, “If we need more computational power, it’s there in Azure HPC, backed by the performance of AMD CPUs. We just make the request, and in a few minutes, we have another node. And we can scale from one node to several on our own.”
Especially important to an automotive company, Rimac can focus on innovation and better meet evolving customer demands now that it can run simulations faster and with greater scale and computing power. In fact, Rimac has reduced its processing time from weeks to days. And when it comes to handling complexity, Krajinović observes, “However complex the model we need to create, we know that we can manage it with Azure HPC. We now produce more highly complex models that simply wouldn’t have been possible on our old infrastructure—they were just too big. And we’re doing iterations for our customers that we couldn’t have done previously.”
Rimac applies the information and insights it gains from faster, more complex simulation models to help it pivot with speed and agility from research and development to market-ready solutions, which amounts to the realization of better products at lower costs. Krajinović states, “Using Azure HPC helps us get accurate engineering information well ahead of production, so we can ensure that no errors exist and that the company won’t lose money fixing late-stage problems. Getting the information faster lets us iterate more and potentially develop an even better product.”
This advanced modeling, testing, and optimization benefits Rimac’s components customers as well. “We conduct our design and production according to the highest, latest standards, modeled with the fastest, smartest hardware and software available,” Krajinović says. “With the power of Azure HPC, we’ll be able to deliver all that value even faster going forward.”
The electric future
The future is bright for Rimac Technology. In addition to continuing to supply components for the Rimac Nevera, the company plans to help develop other Bugatti Rimac cars in the years to come and ultimately become a tier one supplier for the world’s largest automotive OEMs.
The shift from small-scale to large-scale production—making tens of thousands of components—demands capabilities that support better and faster product development at optimized costs. Krajinović concludes, “To position ourselves for much greater volume, we need to be able to quickly and easily conduct simulation optimization. We’re planning on using Azure HPC to do just that. And we’re excited to explore how AI optimization might help us in the future.”
The estimate to get additional physical nodes was 9 to 12 months, and we would have needed 3 more months to install them. Adding a node in an Azure HPC environment, on the other hand, is basically instant.
Ivan Krajinović: Head of Simulations, Rimac Technology