Harnessing the computational capabilities of the graphics processing unit (GPU) is one of the cornerstones of Altair’s mission to empower designers. In this respect, I am pleased to introduce major enhancements to two leading Altair products deployed on GPU technology: Altair AcuSolve™, our advanced computational fluid dynamics (CFD) simulation software, and Thea Render™, a powerful 3D rendering and animation tool. These new enhancements specifically leverage NVIDIA GPUs to solve computational problems up to four-times faster. By utilizing GPU-based CFD solvers on NVIDIA RTX Servers, massive CFD simulations now process in hours rather than days.
External aerodynamics simulation performed with Altair ultraFluidX on the Altair CX-1 concept design, modeled in Altair Inspire Studio
Significant in their own right, these developments also form part of a bigger picture. As we enter a new decade, megatrends such as e-mobility, 5G communication, human augmentation and artificial intelligence (AI) are redefining the boundaries of engineering and creating new industries virtually overnight. Mirroring this, we’re continually seeking to raise expectations of what is possible with simulation.
First and foremost, utilizing GPUs to accelerate numerical simulation delivers significant increases in speed and, therefore, throughput. That means more opportunities to explore and fine-tune designs, make decisions faster based on more accurate results, and consequently, considerably reduce time-to-market. In addition, creativity is enhanced, with more vivid, realistic and accurate rendering and visualization put within easier reach. A little less obvious, customers are free to deploy more energy-efficient computing systems, and establish a flexible, hybrid environment for design and simulation. One that is equally accessible to people using on-site infrastructures or the cloud.
The improvements to AcuSolve help bring these benefits into sharper focus. By implementing its algebraic solver on NVIDIA enterprise GPUs, we’re realizing a six-times faster execution of the most intensive computational operations compared to an equivalent CPU configuration.
Cabin cooling simulation with Altair AcuSolve – model rendered with Thea Render
Validation on RTX Server similarly takes GPU-based CFD solvers ultraFluidX and nanoFluidX to another level. In the words of NVIDIA, “Validated systems deliver unprecedented performance at a fraction of the cost, space and power requirements of traditional CPU-based solutions.”
That means large-scale CFD simulations which previously took days can now be completed overnight and can often be run on a single workstation. The entire simulation workflow can therefore be highly compressed, as shown in the vehicle design example shown below:
The vehicle design shown in the video above was created in Altair Inspire Studio, our industrial design solution that uses Thea Render as the rendering and animation engine. Used in combination with Altair ultraFluidX, it helps designers and aerodynamicists achieve more effective collaboration.
Today, it’s not just design outcomes that are breaking new ground. So are the ways they are being achieved. Major advances in fields such as autonomous mobility are simply too involved to be anything other than the result of teamwork. Successful product development requires ever closer cooperation, collaboration and partnership. Altair’s relationships with leaders like NVIDIA help us push the boundaries of innovation and deliver solutions that inspire our customers.
We firmly believe that an increasingly central role for simulation will help meet these challenges. It’s therefore rather appropriate that our work in conjunction with NVIDIA has such an important role to play. Harnessing the latest advances in GPU technology is far from the full story in terms of enabling our customers to set new standards, but there can be little doubt that its significance will grow as we embrace a future as complex as it is exciting.
View our on-demand sessions from Altair’s Milos Stanic, Yi Chen, and Avadhesh Mittal from the GPU Technology Conference online.