Annons

About Ansys’ och NVIDIA’s Plans to Pioneer the Next Era of Simulation, Analysis and AI

What is more efficient for simulation and analysis work: GPUs (Graphics Processing Units) or CPUs (Central Processing Units)? The question is of some interest, not least when AI is now making a broad entry into product development. Ultimately, the capacity to process heavy data volumes is critical to speed, capability and quality in simulation. According to research (2022) by the consultant Jon Peddie Research (JPR) regarding CAE and, among other things, the use of GPUs, the general opinion among industry players such as Altair, Ansys, Dassault, Hexagon and Siemens Digital Industries Software was that GPUs outperform CPUs by many multiples. In and of itself then depending on the work tasks. Despite this obvious advantage, some engineers were concerned that what is gained in speed may be lost in accuracy. However, according to the interviews conducted by JPR in the report entitled, "Accelerating and Advancing CAE", this is a so far unproven concern. Instead, developers and their customers found that results from GPU-accelerated calculations are as accurate as those performed on CPU-based solvers, but significantly faster. Additionally, several stakeholders believed that the ability to perform more iterations results in better solutions and enables more designers to benefit from simulation earlier in the design process.
This is said as a background to last year's "NVIDIA boom" among the PLM and sub-PLM developers. For example, we have talked about Siemens Digital Industries' Omniverse collaboration and, most recently, SAP's AI partnership around language models (LLM) in NVIDIA AI Foundry and the new NVIDIA NIM Microservices. But there is more to come, this time related to the PLM and CAE domain. Here, Ansys has now announced several new collaborations with NVIDIA, the company that invented the GPU, which it launched back in 1999 - in the form of the integrated graphics processor unit, the Nvidia GeForce 256 DDR. Roughly speaking, this can be considered a starting point for the era of modern AI under the "direction" of NVIDIA. Today, they are also behind much of what happens in industrial digitization, including PLM and ERP.
In connection with the company's GTC event this week, we have for example seen several players in the PLM arena launch collaborations with NVIDIA, not least in relation to NVIDIA Omniverse and digital factory development. One of the more interesting ones is that of Ansys, which announced that it will collaborate with NVIDIA on accelerated computing, advanced 6G communication systems, AI-infused simulation solutions, autonomous vehicles, digital twins, enhanced graphics and visual rendering. "Within the dynamic realm of NVIDIA Omniverse, our visionary customers can drive innovation, bridge virtual and physical realities to shape tomorrow's technologies and help solve the most pressing technological challenges of our time," commented Ajei Gopal, CEO and president of Ansys.

First things first: Ansys is joining the Alliance for OpenUSD (AOUSD) to promote data compatibility for 3D content in products across the entire Ansys portfolio. USD, or Universal Scene Description, is a format originally developed by animation developer Pixar. But OpenUSD is more than a file format – it’s an open source 3D scene description, used for 3D content creation and exchange between tools. Its versatility has made it an industry standard in visual effects, architecture, robotics, manufacturing and more.

On NVIDIA’s part the cooperation will expand its use of Ansys solutions to optimize the hardware, while Ansys will use NVIDIA Blackwell GPUs, Hopper architecture-based GPUs and GB200 Grace Blackwell Superchips to scale up and accelerate existing solutions.


Ansys’ collaboration with NVIDIA is extremely ambitious, with the aim of developing next-generation simulation solutions, powered by, among other things, GPU-accelerated computing and generative AI. The expanded collaboration will fuse cutting-edge technologies to advance 6G technology, supercharge Ansys solvers via NVIDIA GPUs, integrate NVIDIA AI into Ansys software offerings, develop physics-based digital twins, and adapt large language models (LLM) developed with NVIDIA AI foundry services.

Alreday in January, Ansys announced that Ansys AVxcelerate Sensors will be available within NVIDIA DRIVE Sim, a scenario-based AV simulator powered by NVIDIA Omniverse, a platform for developing Universal Scene Description (OpenUSD) applications for industrial digitization. The integration will provide users with access to high-hostile sensor simulation outputs. These outputs are generated using Ansys AVxcelerate sensors for ADAS/AV system training and validation.

Strengthened compatibility and sharpened graphics for smart factories
Thus, Ansys recently joined AOUSD to strengthen data compatibility and deliver improved graphics and visual rendering to its portfolio. Ansys has already connected Ansys AVxcelerate Autonomy to NVIDIA DRIVE Sim powered by the NVIDIA Omniverse platform and plans to explore further integrations across the portfolio, including Ansys STK, Ansys LS-DYNA, Ansys Fluent and Ansys Perceive EM. This seamless interoperability will empower users to tackle a wide range of challenges, from factory-level to planetary-level simulations.

CAPTION Ansys Perceive EM in NVIDIA Omniverse models 5G/6G antenna signals from moving vehicles in Denver.

In this, digital twins play a leading role.

”That’s right,” says NVIDIA CEO Jensen Huang. “Everything manufactured will have digital twins – and the world’s designers and engineers in the heavy industrial markets rely on Ansys as their simulation engine. We are partnering with Ansys to bring accelerated computing and generative AI to these massive workloads, and to extend Ansys’ leading physics simulation tools with NVIDIA Omniverse digitization technology.”

Four priority areas
In addition to Omniverse integrations, the collaboration will prioritize progress in four areas:

Accelerated Computing: Ansys, working with NVIDIA, enables customers in various industries to shorten design cycles and deliver increasingly complex products by advancing numerical research in high-performance computing. Ansys leverages NVIDIA H100 Tensor Core GPUs to power multiple simulation solutions and prioritizes NVIDIA Blackwell-based processors and NVIDIA Grace Hopper Superchips for products across the Ansys portfolio, including Ansys Fluent, Ansys LS-Dyna, and Ansys electronics and semiconductor products. At the same time, NVIDIA leverages Ansys technology, including semiconductor tools, to enrich virtual models and data center design, ultimately leading to accelerated Ansys solver performance.

6G Telecommunications: Ansys is among the first to use the NVIDIA 6G Research Cloud platform, providing researchers with a comprehensive suite for developing AI for radio access network (RAN) technology. Ansys Perceive EM, a new solver powered by Ansys HFSS, is built on the NVIDIA 6G Research Cloud, which is designed to accelerate the development of 6G technology. Perceive EM’s synthetic data-on-demand revolutionizes the 6G system’s digital twins with heightened predictive accuracy, which can assess how real-world conditions affect wireless network performance. Perceive EM will also be available within the Ansys Academic and Ansys Startup Program, which have reached 2,900+ universities and 2,100+ startups, respectively.

AI-enhanced simulation: Ansys is exploring the NVIDIA Modulus framework for physics-based machine learning to further strengthen its software offering with the latest AI technologies. The work aims to deliver improved functionality within the Ansys AI+ product family, such as more efficient optimization, sensitivity analyzes and robust designs.

AI Foundry: Ansys is exploring the introduction of NVIDIA AI Foundry to advance LLM development, which promotes the democratization of simulation by simplifying setup and use. Future LLMs tailored to Ansys solutions offer the potential to provide virtual expert assistance that opens the door to new users and creates new simulation use cases. Ansys intends to leverage the NVIDIA NeMo platform, which offers a set of tools that make it easier, more cost-effective and faster to develop generative AI functions.

Print Friendly, PDF & Email

Success Stories

Industriellt

Success Stories

Intressant på PLM TV News

PLM TV News

PLM TV News

PLM TV News

PLM TV News

PLM TV News

Aktuell ANALYS

Aktuell Analys

Aktuell Analys

3D-printing

Block title