Highlights:

  • Nvidia Corporation has formed a partnership with Rescale Inc., a high-performance computing company, to simplify the operation of artificial intelligence applications on public cloud infrastructure.
  • Rescale’s technology automates a significant portion of the human labor required to deploy HPC software on cloud infrastructure.

Nvidia Corporation has partnered with Rescale Inc., a high-performance computing company, to simplify the operation of Artificial Intelligence (AI) applications on public cloud infrastructure.

This alliance announced today has numerous components. Rescale will combine its technology with several AI development and deployment tools from Nvidia. In addition, the business is releasing software that automatically optimizes the cost and performance of an enterprise’s AI application environment.

San Francisco-based Rescale has received about USD 150 million in financing from Nvidia, Microsoft Corporation, and other notable investors. It offers a software platform that facilitates the execution of High-Performance Computing (HPC) applications in the cloud. HPC applications are hardware-intensive programs that assist research and engineering projects, such as physics simulators.

Rescale’s technology automates a significant portion of the human labor required to deploy HPC software on cloud infrastructure. It also assists with related responsibilities, such as application security.

Edward Hsu, Resale’s chief product officer, said, “We’re moving from what I’d call intuition-based engineering to a more scientific model.”

Hsu offered several illustrations of how the technique operates. For example, a corporation building an airfoil was previously required to create designs and test them in a wind tunnel, a lengthy and expensive operation. On Rescale, he added, 20 to 30 airfoil designs may be simulated without a wind tunnel. Now, with the additional capabilities of Nvidia, they can utilize AI and Machine Learning (ML) to identify which strategies are optimal for the task at hand.

Rescale offers a catalog of pre-packaged HPC applications as part of its features. Customers may install the apps without the manual work that is often required. Rescale is adding support for many of the chipmaker’s AI tools to its software catalog through its newly announced relationship with Nvidia.

Rescale will add support for the Nvidia AI Enterprise platform as its initial offering. It is a collection of more than a dozen software tools to facilitate the construction, deployment, and maintenance of neural networks.

Several of the Nvidia AI Enterprise platform’s solutions aim to simplify the management of the cloud infrastructure on which an organization’s AI software operates. Others can improve neural networks to use Nvidia chips more efficiently. Nvidia has incorporated software building blocks that make it easier for developers to design AI apps for use cases such as robotics.

Rescale will add support for more software tools to its platform, additional to Nvidia AI Enterprise.

The Nvidia Base Command Platform is one of the tools that provide researchers an interface to manage the training process of a new neural network. Rescale additionally adds support for Nvidia Modulus, another software offering. The offering facilitates the development of neural networks for research and engineering applications involving physics calculations.

Nvidia founder and Chief Executive Officer Jensen Huang said, “Fusing principled and data-driven methods, physics-ML AI models let us explore our design space at speeds and scales many orders of magnitude greater than ever. Rescale is at the intersection of these major trends. Nvidia’s accelerated and AI computing platform perfectly complements Rescale to advance industrial, scientific computing.”

Rescale today unveiled the Compute Recommendation Engine, or CRE, a new tool in addition to the expanded support for Nvidia software. The solution is made to assist businesses in streamlining the cloud environments in which they operate for AI and high-performance computing applications.

According to Rescale, CRE may cut hardware expenses by optimizing the infrastructure resources added to a cloud environment. Alternatively, the tool can optimize performance by configuring infrastructure resources. Organizations may tailor how CRE balances cost-effectiveness and performance to meet their needs.

Rescale founder and CEO Joris Poort said, “Our unique built-for-the-cloud approach enables us to deliver optimal accelerated computational performance on-demand for any workload on any cloud worldwide, and our collaboration with Nvidia will bring powerful new AI capabilities to our industrial HPC customers.”

Rescale asserts that enterprises utilize their platform for use cases like medical research, chip design, and aircraft development. The startup’s new CRE infrastructure optimization tool was built using metadata collected from over 100 million high-performance computing applications.