Domino Data Lab launches fully-managed MLOps service with Nvidia

San Francisco, California-based Domino Data Lab, a provider of MLOps solutions, has announced a fully managed offering with Tata Consultancy Services and Nvidia to help enterprises unite their analytics and AI workloads with high-performance computing in a single environment.

Unveiled at the ongoing Nvidia GTC conference, the solution leverages Domino’s MLOps platform and runs high-performance computing and data science workloads on Nvidia DGX systems, all in the TCS enterprise cloud.

Converged solution for MLOps

The MLOps offering works as a single, converged end-to-end solution for training AI, ML, and deep learning models using Domino and Nvidia DGX systems in the same heterogeneous compute environment as HPC simulation workloads. This way, data science leaders can use the output data from CPU-accelerated containerized simulation workloads to train Nvidia GPU-accelerated ML models, or vice versa, without having to move data across two traditionally siloed environments.

The flexible procurement and deployment options also ensure that data science leaders can use the solution to track project status across teams, while allowing IT teams to track infrastructure utilization for capacity planning.

“The most difficult challenges in industries like life sciences and manufacturing can be solved by efficiently leveraging the proliferation of data from connected devices, and by converging simulation, analytics, and AI/ML workloads in a single environment,” Dr. Revati Kulkarni, technology head for HPC at TCS AI group, said.

“TCS’s HPC A3 solution uses Domino’s capabilities to seamlessly manage heterogeneous compute across complex datasets and helps customers accelerate their transformation journey,” she added.

Nick Elprin, the CEO and co-founder of Domino Data Lab, noted that the development will not only help the company’s customers accelerate breakthrough research but also increase the productivity of their data science teams.

Domino platform to integrate with Nvidia AI Enterprise

The engagement with Nvidia for the new solution comes as Domino’s MLOps platform inches closer to the integration with Nvidia AI Enterprise, an end-to-end, cloud-native suite of AI and data analytics software optimized for the Nvidia EGX platform, running on mainstream Nvidia-certified systems from OEM hardware providers and VMware vSphere. Domino said that validation is underway for the integration of the platform.

Just last month, the company had also raised $100 million in a series F round of funding. It said that the capital from the round would largely go toward product development and expansion of the MLOps platform to grow customers worldwide.

According to a study from Cognilytica, the MLOps market could grow from $350 million in 2019 to $4 billion by 2025.

Source