Rescale has integrated the Nvidia GPU Cloud (NGC) into its library of containers for high-performance computing (HPC) apps. This will make it easier for enterprises to build digital twins and enable digital transformation as a foundational building block for the industrial metaverse. 

The partnership with Nvidia adds over 150 additional containerized artificial intelligence (AI) and HPC applications and hundreds of pretrained models optimized for Nvidia GPUs. This joins more than 900 other applications already pre-integrated into the Rescale platform. The companies are also working on adding integration to the Nvidia Base Command AI training platform early next year. 

Optimized for digital twins

Manufacturers are increasingly using computational engineering for product innovation. The Rescale integration with Nvidia will simplify the process of designing and testing digital twins of products. This automation will help researchers assess multiple iterations of digital twins representing different assumptions about the product to identify and mitigate more kinds of problems. 

Rescale chief product officer Edward Hsu told VentureBeat, “Tackling digital twins and the industrial metaverse require tremendous computing power best supported by specialized computing architectures and software codes optimized to take advantage of them.”  

This is true for workloads ranging from AI/ML (machine learning) to physics-based simulations that predict how products would perform in the real world. Rescale customers can now use a single platform to do AI-assisted computational engineering, taking advantage of the latest Nvidia architectures and performance-optimized software tools on any cloud.

Rescale had previously supported integrations with specific AI/ML offerings such as Modulus, PyTorch, and TensorFlow, as well as NGC-optimized software. The new integration will improve workflows for over one hundred prebuilt containers out of the box. Hsu said this effectively increases the number of AI/ML-related workflows that can be automated on the platform by 10 times. 

“What’s unique about NGC is that these applications are tuned, tested and optimized by Nvidia,” Hsu said. 

Digital industry ecosystem

The scientific and engineering community is slowly migrating away from relatively cumbersome software development processes compared to other kinds of enterprise software. One challenge is that these apps typically require considerably more customization and must be deployed with dependencies, making it harder to deploy new versions quickly. Gartner observed that pioneering enterprises are increasingly adopting cloud infrastructure for engineering workloads, which could help drive the HPC market to $55 billion by 2024. 

The new partnership between Rescale and Nvidia will allow enterprises to connect workflows between Rescale’s existing catalog of engineering and scientific containers, Nvidia’s extensive NGC offerings, and enterprises’ standard containers of their own models and supporting software. This new containerized approach to engineering software means teams can specify the software libraries and configurations that reflect industry best practices. 

The recent Nvidia and Siemens partnership is an ambitious effort to bring together physics-based digital models and real-time AI. Rescale’s announcement with Nvidia enhances this partnership, as accelerated computing combined with high-performance computing is the foundation that powers these use cases.

For example, enterprises can take advantage of Nvidia’s work on Modulus, which uses AI to speed up physics simulations hundreds or thousands of times. Siemens estimates that integrating physics and AI models could help save the power industry $1.7 billion in reduced turbine maintenance. The partnership could also make it easier for companies to integrate other apps that work on these tools. 

The rubber meets the digital road

Companies like Hankook Tire in South Korea are already taking advantage of the new integration to accelerate R&D for new tire designs. The partnership allows Hankook to combine its proprietary algorithms with off-the-shelf components from Nvidia and other Rescale partners to accelerate the design, optimization and testing of new product iterations. 

Hankook Tire is developing digital twins of their tires and virtually testing them through various conditions with their own custom-made models and codes, aiming to develop new products more quickly and improve performance. 

There is high demand for the latest GPU architectures in the cloud, and capacity in specific regions may not meet demands. With Rescale, the company can automate its proprietary engineering software across multicloud operations by load-balancing across different hardware architectures and geographic locations. Through automated optimization, Hankook Tire can prioritize running the latest Nvidia GPU solutions available in Korea and switch to other new Nvidia architectures and global cloud capacities seamlessly, without impacting any use by R&D engineers or manual work by IT teams.

This recent partnership demonstrates the power of improved automation, high-performance computing and digital twins to drive industrial digital transformation efforts. “As organizations look to do AI-assisted engineering, these capabilities allow companies to have a single platform to tackle the software, hardware, and workflow challenges needed to design and deliver engineered products,” Hsu said.