Nvidia and Databricks Expand Collaboration to Boost Enterprise AI Development


Nvidia and data and artificial intelligence (AI) company Databricks have deepened their collaboration.

The partnership aims to optimize data and AI workloads on the Databricks Data Intelligence Platform utilizing Nvidia’s accelerated computing and software, the companies said in a Tuesday (March 19) press release.

“From analytics use cases through AI, Nvidia has already powered our foundational model initiatives, and with our mutual work on query acceleration, we’ll be able to demonstrate value for more enterprises,” Ali Ghodsi, co-founder and CEO at Databricks, said in the release.

Jensen Huang, founder and CEO of Nvidia, added: “By accelerating data processing, Nvidia and Databricks can supercharge AI development and deployment for enterprises seeking greater insights and better outcomes with more efficiency.”

The collaboration’s focus includes GPU acceleration for generative AI solutions, according to the release. Organizations are increasingly adopting Databricks’ platform to build and customize generative AI models. Databricks Mosaic AI and Nvidia are working together on model training and inference to advance generative AI models’ building and deployment. Databricks leverages Nvidia’s H100 Tensor Core GPUs for this purpose, providing an efficient and scalable platform for these tasks.

Furthermore, Databricks plans to develop native support for Nvidia accelerated computing in its next-generation vectorized query engine, Photon, the release said. This development will enhance speed and efficiency for customers’ data warehousing and analytics workloads. Photon powers Databricks SQL, the company’s serverless data warehouse, and this development builds on the growing use of GPUs for query processing on Databricks’ platform.

Databricks also offers comprehensive support for machine learning (ML) and deep learning workloads, per the release. Databricks Machine Learning delivers pre-built deep learning infrastructure with Nvidia GPUs, and the Databricks Runtime for ML includes pre-configured GPU support.

Over 10,000 organizations worldwide, including over 50% of the Fortune 500, rely on the Databricks Data Intelligence Platform to unify and democratize data, analytics and AI, according to the release.

Companies want to use their data more effectively, Junta Nakai, global head of financial services at Databricks, told PYMNTS in an interview posted in February.

To do so, they need to modernize their tech stack, democratize access to data throughout their company, and then transform their company, Nakai said.

“That’s what we do with our technology,” Nakai said. “And simply put, it stores all your data in a single place, and do all the things that you do with your data come from that.”