Data Center-Ready MLOps: Domino now validated for NVIDIA AI Enterprise
By Thomas Robinson2022-01-196 min read
Today, Domino Data Lab is announcing the validation of Domino’s enterprise MLOps platform integrated with NVIDIA AI Enterprise and VMware vSphere with Tanzu, helping enterprises worldwide manage data science and AI workloads using mainstream servers. Domino is the first MLOps software NVIDIA has validated with the NVIDIA AI Enterprise software suite to provide the highest performance and compatibility on NVIDIA-certified Systems from leading server providers.
GPU acceleration brings promise, but not without challenges
From natural language processing for model-driven policy approvals to image classification in biotechnology to supply chain risk and anomaly detection in advanced manufacturing, the competitive advantage from GPU-accelerated model training is clear. The intensity of data science workloads, particularly for model development and deep learning, results in varying compute infrastructure requirements, with peak utilization orders of magnitude greater than average utilization.
However, enterprises frequently struggle with AI adoption because of the complexity of integrating these AI solutions with existing enterprise infrastructure. This is compounded by data scientists’ need for flexibility in tooling across the data science lifecycle, resulting in siloed environments between data scientists and IT.
Breaking down silos between data scientists and IT
Domino Data Lab’s enterprise MLOps platform is now validated and integrated with NVIDIA AI Enterprise, an end-to-end, enterprise platform software suite optimized for AI workloads. This joint solution between NVIDIA, VMware, and Domino Data Lab marks a new phase in the rapid innovation, scale, and expansion of AI by bringing powerful AI solutions into enterprise data centers on mainstream servers.
On infrastructure already familiar to IT, data science teams can now easily complete projects such as training an image recognition model using TensorFlow, performing textual analysis with NVIDIA RAPIDS, or deploy an intelligent chatbot with Triton Inference Server. Breaking down silos across data science and IT opens the door for even more enterprises to build a competitive advantage by embedding models throughout the business.
Data Center-Ready Enterprise MLOps
NVIDIA AI Enterprise is optimized and certified for VMware vSphere® with Tanzu, the industry’s leading virtualization platform. This provides AI developer frameworks and tools backed by NVIDIA support, with kubernetes operators, drivers, and enterprise-supported containers necessary for data science and AI workloads.
Domino’s Enterprise MLOps platform works on top of VMware vSphere, integrating NVIDIA AI Enterprise containers with the Kubernetes layer provided by VMware Tanzu. This provides an end-to-end platform for the research, development, and deployment of data science and AI workloads. The NVIDIA AI Enterprise validation means this entire solution can be deployed on NVIDIA-Certified systems from server vendors and integrators like HPE, Dell, Lenovo, Cisco, and Hitachi Vantara.
Upskilling data scientists with the right tools and infrastructure
Domino uniquely enhances the NVIDIA AI Enterprise software suite by providing an end-to-end enterprise MLOps platform, a system-of-record for data science work. A single click can spin up research environments with optimized images for NVIDIA RAPIDS, TensorFlow, and PyTorch frameworks on NVIDIA-Certified Systems with data, code, and prior work automatically loaded alongside access to NVIDIA GPUs. This automation allows data scientists to focus on innovative research and model development instead of burdensome devops work.
Similarly, Domino handles all the model hosting and inference, so once a data scientist has created a model, it can easily be hosted and maintained in the environment.
Simplified management and reduced support burden for IT
For IT, this integration dramatically simplifies support requests from data scientists by providing a virtualized, cloud-like “data science platform” all hosted within the familiar resilient VMware vSphere enterprise infrastructure environment running non-AI applications. IT can quickly provision pre-configured containers on-demand with allocated GPU resources, reducing the number of help-desk requests from data scientists for infrastructure and licenses. Domino provides a self-service solution for data scientists - with reporting and governance for IT.
With the addition of the NVIDIA AI Enterprise validation, Domino has many different products and offerings with NVIDIA and the NVIDIA Partner Network (NPN). We continue to expand our NVIDIA DGX-Ready Software partnership, have recently announced an NVIDIA-based managed service offering with TCS, and continue to support NVIDIA accelerated computing on the major cloud providers. Our mission is to unleash the power of data science for the world’s most important enterprises - democratizing GPU access is an important pillar in that mission.
- Join Domino on January 26th for a virtual event to learn about Domino 5.0, a major step forward in our mission to help businesses unleash the power of data science.
- Learn more about Domino and NVIDIA's work together, including customer use cases and technical resources.
- Download the Domino & NVIDIA AI Enterprise deployment guide for technical details.
Thomas Robinson is the VP of Strategic Partnerships and Corporate development at Domino, where he's responsible for building Domino's partner ecosystem, developing offerings providing differentiated value to partners. He previously acted as Domino's chief people officer, responsible for building an organization to unleash data science to address the world's most important challenges. Prior to Domino, Thomas worked at Bridgewater Associates, driving strategic transformation efforts, first as a director in Bridgewater's Core Technology Department to define the next generation of enterprise architecture, and then as a general manager focused on recruiting and retaining technical talent.
Subscribe to the Domino Newsletter
Receive data science tips and tutorials from leading Data Science leaders, right to your inbox.