Domino Data Lab Debuts New Solutions with NVIDIA to Enhance the Productivity of Data Scientists
April 13, 2021
Allowing the Full Utilization of NVIDIA solutions, platforms, and technologies, the Domino Enterprise MLOps Platform Increases Capacity of Data Scientists
SAN FRANCISCO – April 12, 2021 – Domino Data Lab, provider of the leading Enterprise MLOps platform trusted by over 20% of the Fortune 100, today announced at NVIDIA’s GTC conference a series of new integrated solutions and product enhancements with NVIDIA, some of which are available now and others will be made available in the coming months. These new enhancements allow data scientists and data engineers the ability to deploy the industry’s most powerful and innovative solution to enhance productivity and positively impact business outcomes.
“Domino’s enterprise MLOps platform accelerates research, speeds model deployment, and increases collaboration for code-first data science teams at scale,” said Nick Elprin, CEO and co-founder at Domino Data Lab. “By having the opportunity to work closely with NVIDIA, we are able to help data scientists fully realize the scope and possibilities of NVIDIA technologies, through high-performing solutions that deliver real business benefits to organizations - something we’ll be showcasing at NVIDIA GTC.”
The Domino Enterprise MLOps platform centralizes data science work and infrastructure across the enterprise for collaboratively building, training, deploying, and managing models faster and more efficiently. With Domino, data scientists can innovate faster, teams can reuse work and collaborate more efficiently, and IT teams can manage and govern infrastructure. To support this, Domino is announcing at GTC a variety of solutions developed in collaboration with the NVIDIA team. These solutions include:
Domino Available for NetApp ONTAP AI Integrated Solution
To streamline the adoption of AI Infrastructure, NVIDIA teamed with NetApp to introduce NetApp ONTAP AI integrated solution, a turnkey offer consisting of pre-integrated compute, networking, and storage components. But one hurdle has remained - enhancing data science productivity with software that streamlines the workflow while maximizing infrastructure utilization. To manage and operate the infrastructure you’ve invested in, Domino’s Data Science platform is now available alongside the NVIDIA and NetApp hardware. Domino has been tested and validated to run on the packaged offering and is available for purchase through the NVIDIA Partner network with special packaging and pricing for NetApp ONTAP AI bundles.
Expanded Multi-Node Cluster Offerings
Many deep learning and AI training jobs require more than a single NVIDIA GPU or single multi-GPU machine. Setting up a multi-node cluster can be difficult, so many teams decide to leave these resources in place, even if that means they’ll often sit idle when there are no complex experiments to be run. Dedicated resources and low utilization are eliminated with Domino. The platform automatically creates and manages multi-node clusters, and releases clusters when training is done. Domino currently supports ephemeral clusters using Apache Spark and Ray, and will be adding support for Dask in a product release coming this Fall.
NVIDIA MIG Support for Maximizing GPU Usage
With Domino’s new support for NVIDIA’s Multi-Instance GPU (MIG) technology, administrators can easily divide a single NVIDIA DGX A100 GPU into multiple instances or partitions to support a variety of users and their use cases. For example, a single NVIDIA A100 GPU can allocate seven MIG devices. This allows 7x the number of data scientists to run a Jupyter notebook attached to a single GPU versus without MIG. By providing various sized compute options, more data scientists can use the system simultaneously, and companies get the maximum benefit from their GPU investment. With MIG technology on NVIDIA DGX A100 systems, Domino can take this even further, allowing up to 56 concurrent notebooks or hosted models, each with a partitioned GPU. Support for MIG will be added in our September release.
NVIDIA NGC Containers Available for Domino
The NVIDIA NGC catalog empowers researchers, data scientists, and developers with performance-engineered, GPU-optimized hub of AI and HPC containers, pre-trained models, industry SDKs and Helm Charts. The NGC catalog simplifies and accelerates end-to-end workflows, enabling users to focus on building lean models, producing optimal solutions, and gathering faster insights. Today, Domino enables customers to create framework-specific custom work environments based on NGC containers, including RAPIDS, TensorFlow, PyTorch, CUDA, and more.
“Domino’s expertise in managing and scaling data science with discipline and maturity, along with its history of driving innovation, are just a few standout reasons why they are highly valued data science partner,” said Charlie Boyle, vice president and general manager, DGX systems at NVIDIA. “The DGX team looks forward to continuing to work collaboratively with Domino on our quest to offer the most powerful deep learning, artificial intelligence, and accelerated analytics solutions on the market.”
Domino is a platinum sponsor of NVIDIA GTC, which runs today, April 12 through Friday, April 16. Learn more about NVIDIA and Domino Data Lab sessions at GTC, including industry-specific enterprise use cases on topics ranging from accelerating model deployment to embedding data science across an organization.
About Domino Data Lab
Domino Data Lab empowers data science teams with the leading, open Enterprise MLOps platform that enables enterprises to manage and scale data science with discipline and maturity. Model-driven companies including Allstate, Dell Technologies, and Bayer use Domino as a data science system of record to accelerate breakthrough research, increase collaboration, and rapidly deliver high-impact models. Founded in 2013 and based in San Francisco, Domino is backed by Sequoia Capital, Coatue, Bloomberg Beta, Dell Technologies Capital, Highland Capital Partners, and Zetta Venture Partners. For more information, visit dominodatalab.com.