San Francisco, March 18, 2024 /CNW/ — data bricka data and AI company, today announced expanded collaboration and deeper technology integration with NVIDIA during the company's flagship period. GTC 2024 meeting. Databricks and NVIDIA are working together to optimize data and AI workloads on the Databricks Data Intelligence Platform. This collaboration builds on NVIDIA's recent participation in Databricks. Series I funding round.
“We are excited to evolve our partnership with NVIDIA and bring more value to our customers through the evolution of Databricks workloads with NVIDIA accelerated compute and software,” he said. Ali Godoshi, co-founder and CEO of Databricks. “From analytics use cases to AI, NVIDIA is already ramping up our work on our foundational models, and our mutual work on query acceleration will allow us to demonstrate value to more enterprises.”
“Each company's proprietary data is a critical asset for creating intelligence for the AI era,” said Jensen Huang, Founder and CEO of NVIDIA. “By accelerating data processing, NVIDIA and Databricks can significantly enhance AI development and deployment for companies seeking greater efficiency, better insights and better results.”
GPU acceleration for end-to-end generative AI solutions
Organizations are rapidly adopting Databricks' data intelligence platform to build and customize generative AI solutions that are trained on data and tailored to their business and domain. Databricks Mosaic AI and NVIDIA are collaborating on model training and inference to advance the state of building and deploying generative AI models on Databricks' end-to-end platform. Databricks provides a comprehensive set of tools to build, test, and deploy generative AI solutions with uncompromising control and governance over both your data and models.
For training generative AI models, Databricks Mosaic AI relies on NVIDIA H100 Tensor Core GPUs, which are optimized for LLM development. Mosaic AI can leverage the power of NVIDIA accelerated computing to provide an efficient and scalable platform for customizing LLM for customers.
For model deployment, Databricks leverages NVIDIA accelerated compute and software across the stack. The key components of Databricks' Mosaic AI model offering are: NVIDIA TensorRT-LLM Software that provides state-of-the-art performance and ensures that your solutions are cost-effective, scalable, and performant.Mosaic AI TensorRT-LLM is a launch partner and enjoys close technical collaboration with the NVIDIA team.
Photon with NVIDIA accelerated computing accelerates query performance
Databricks plans to develop native support for NVIDIA-accelerated computing into its next-generation vectorized query engine, Photon, to improve the speed and efficiency of customers' data warehousing and analytics workloads. Photon powers Databricks SQL, his Databricks serverless data warehouse with industry-leading price performance and TCO. It advances and builds on the growth of Databricks customers using GPUs to process data queries.
Databricks for machine learning and deep learning
Machine learning (ML) and deep learning on Databricks are important workloads. Databricks Machine Learning provides prebuilt deep learning infrastructure that includes NVIDIA GPUs, and the Databricks runtime for ML includes preconfigured GPU support, including drivers and libraries. These tools allow a user to start using the right NVIDIA infrastructure right away and keep the environment consistent across users. Databricks supports NVIDIA Tensor Core GPUs on all three major clouds, enabling high-performance single-node and distributed training of ML workloads.
The companies plan to further build on the momentum of their data intelligence platform, enabling more organizations to create next-generation data and AI applications with quality, speed, and agility.
About data bricks
Databricks is a data and AI company. More than 10,000 organizations around the world, including Comcast, Condé Nast, Grammarly, and more than 50% of the Fortune 500, rely on the Databricks data intelligence platform to unify and democratize data, analytics, and AI. Databricks' headquarters is San Franciscohas offices around the world and was founded by Apache Spark™, the original creator of Lakehouse. delta lake, MLflow.For more information, follow Databricks linkedin, X and Facebook.
Press contact:
press@databricks.com
View original content to download multimedia: https://www.prnewswire.com/news-releases/databricks-and-nvidia-deepen-collaboration-to-accelerate-data-and-ai- workloads-with-the-data-intelligence-platform-302092139.html
Source Databricks
View original content to download multimedia: http://www.newswire.ca/en/releases/archive/March2024/18/c0515.html