Domino Data Lab Teams With Nvidia, NetApp To Enable MLOps Across Hybrid Environments

New integrations between Domino Data’s MLOps software, Nvidia high-performance GPUs and NetApp data management solutions, along with a new reference architecture, will speed AI/ML workloads across hybrid and multi-cloud environments, the companies say.

MLOps platform developer Domino Data Lab has teamed up with chipmaker Nvidia and data management and storage system supplier NetApp to develop new solutions and a new reference architecture that will facilitate machine learning and data science workloads running across hybrid and multi-cloud systems.

In an announcement Tuesday, made during Nvidia’s GTC developer conference, Domino Data Lab, Nvidia and NetApp unveiled new integrations between Domino Data Lab’s Nexus hybrid and multi-cloud MLOps architecture, Nvidia GPUs and NetApp data management and storage systems.

The companies say the new integrated solutions and reference architecture will allow businesses and organizations to more easily run and manage AI/ML and data science workloads either within on-premises data centers or on the Amazon Web Services cloud platform without refactoring them.

[Related: Domino Data Lab Launches Inaugural Partner Program Targeting Service, Technology Providers]

“It’s all of us working together with all of our hybrid capabilities to provide a hybrid, multi-cloud solution for customers,” said Thomas Robinson, vice president of strategic partnerships and corporate development at Domino Data Lab, in an interview with CRN. “We’re hearing a lot from customers that hybrid and multi-cloud is the future of machine learning and data science work.”

At the center of the newly unveiled solution is Domino Data Lab’s Nexus hybrid and multi-cloud MLOps architecture that makes it possible to rapidly scale, control and orchestrate data science workloads across different compute clusters – whether they be in different geographic regions, on premises or across multiple clouds.

Nvidia is a development partner for Nexus, which will run machine learning workloads on Nvidia GPUs. Unveiled in June, Nexus is slated to go into private beta later this month and be generally available in early 2023.

NetApp’s data management and storage systems come into play because a lot of the data being used for AI, machine learning and data science tasks is being moved from older file systems and Hadoop data stores to next-generation databases and Spark-based data lakes, said Phil Brotherton, NetApp vice president of solutions, in an interview with CRN.

“Everyone’s trying to adopt AI,” Brotherton said. “We see AI as a use case that’s just going to keep rolling. And where NetApp gets involved is typically in step one in getting the data collected and processed in a way that it can be moved correctly.”

NetApp, which develops AI data management systems, will validate Domino Nexus as a new solution that supports the Domino Enterprise MLOps Platform on Amazon FSx for NetApp ONTAP. The new AWS Managed Service solution will support evolving hybrid workload requirements and simplify the deployment and management of large-scale applications in hybrid real-time environments, according to the companies

Through the three-way collaboration Domino Nexus will provide a holistic view of enterprise data science workloads across multiple regions and environments, including the new AMS solution and existing on-premises NetApp ONTAP AI systems, according to the companies.

NetApp Cloud Manager will serve as a control plane to set up data replication across multiple environments via SnapMirror, an approach that automates data science processes and accelerates AI workload deployment, the companies said.

“The big picture for everybody, for all customers and channel partners, is that it’s becoming quite real that you can use AI where you want to use it – on premises or in the cloud – and use it in the best possible ways,” NetApp’s Brotherton said.

“I tell all of my partners that AI and modern data analytics is going to be huge on premises,” Brotherton added. “This is all wide open right now for channel partners. This is early days [for] a huge opportunity.”

Domino Data Lab and Nvidia have also developed a reference architecture for integrating Domino Data Lab’s MLOps with on-premises Nvidia DGX systems. The reference architecture provides a full-stack solution guide for running Nexus on large-scale, multi-node compute systems based on Nvidia DGX, Robinson said.

The reference architecture also will support partners as they establish AI centers of excellence and build end-to-end AI platforms using Domino Data Lab and Nvidia technologies for their customers. “Solution providers can take this blueprint to a customer and say: ‘We know what the stack looks like. It’s validated. We have a way to implement it,’” Robinson said.

Mark III Systems, a Houston-based IT transformation and AI consultant and solution provider, worked with Domino Data Lab and Nvidia to develop the reference architecture. The company partners with Nvidia and Domino Data Lab to help build IT stacks for AI, ML and data science initiatives and “bring it all together using our team,” said Andy Lin, CTO and vice president of strategy and innovation at Mark III, in an interview with CRN.

Lin added that Domino Data Lab’s MLOps technology is “the bridge” that will allow businesses to manage AI and data science initiatives across on-premises, hybrid and multi-cloud environments. “This partnership with Domino specifically, and in general building our AI practice, has been a game-changer for us as an organization.”

“I think hybrid MLOps, quite frankly, is really the wave of the future,” Lin said. “The hybrid approach is here to stay.”

Domino Data and Nvidia already have a multi-faceted alliance that includes product integration and joint sales efforts. Nvidia is an investor in Domino Data, participating in Domino Data’s $100 million Series F funding round in October 2021. Also in 2021 Domino Data and Nvidia allied with global IT services giant Tata Consulting Services with TCS incorporating Domino’s data science and machine learning software as part of a managed services platform based on Nvidia DGX systems. Domino Data also recently joined the Nvidia AI Accelerated program for software and solution partners.