Ansys User Conference 2023
14. 6. 2023
14. 6. 2023
The event was held in the beautiful surroundings of the OREA Resort Horal in Spindleruv Mlyn and focused on news and other topics from the Ansys software, which is one of the world’s top in the field of technical simulations and physical calculations.
During our lecture session, we presented the benefits of multi-GPU acceleration of these computations using the latest NVIDIA accelerators and also presented the results of our partners’ testing using Ansys software.
Thanks to the high number of computational cores, multi-GPU acceleration brings significant speedup in computations and simulations. The use of multiple GPUs within a single system enables the processing of extremely demanding tasks and large mathematical models.
The GPU accelerator can do the same type of computation not only faster, but more importantly, more efficiently and using less power. This results in lower cooling requirements for the entire system and even greater energy savings.
With faster job processing and better energy efficiency, less hardware is needed to do the same amount of computation, which not only significantly reduces the initial purchase investment, but also saves space in your datacenter.
Ansys has made a significant focus in recent years on supporting compute acceleration on GPU cards. Probably the best optimization for real-time GPU computing is currently its Ansys Fluent software. It is a professional flow simulation software known for its advanced physical modelling capabilities and superior accuracy.
Its latest version (2023 R1) already includes an advanced multi-GPU solver that speeds up and streamlines a large number of calculations, whether it is simulating airflow around wind turbines or drone rotors or thermal heating of batteries and electric motors. With each release, the Fluent multi-GPU solver improves and delivers better and better results even on the same hardware.
In preparation for the conference, together with our partners TechSoft and New Technologies – Research Center (NTC) from the University of West Bohemia in Pilsen, we managed to test selected GPU accelerators from our NVIDIA demo warehouse.
It features the latest and most powerful technologies – the NVIDIA DGX Station A100 workstation and NVIDIA H100, A100, A30, A16, A2 or e.g. RTX 6000 Ada Lovelace.
More details, measurement results and comments can be found in the downloadable presentations below.
To test the performance and speed of deploying ML and AI applications, we not only have NVIDIA DGX Station A100 systems and NVIDIA A100 accelerators as part of the NVIDIA Tesla Test Drive program, but also demo licenses for GPU virtualization (vGPU) and NVIDIA AI Enterprise. If you are interested in our testing offer, please fill out this form.