Solution guide
Page 2
SERVICE DESCRIPTION HPE GreenLake for ML Ops is an on-premises end-to-end solution for data science that includes hardware, software, and services. It brings the cloud to the data center providing data scientists and their Dev Ops counterparts with an enterprise-scale, data science service that keeps data on premises. The HPE GreenLake ML Ops service makes it easier and faster to deploy AI/ML workloads on premises with a cloud experience: self-service, pay-per-use, scalable, and managed for you. The service platform integrates hardware, software, and service to enable customers to build, train and deploy ML workloads quickly and at scale. The offering supports the end-to-end data science lifecycle in a flexible design that provides the data scientist a single environment to work. By removing the operational heavy lifting from each step of the ML process, HPE GreenLake for ML Ops provides the required ML tools, making it easier to deploy high-quality ML models and allowing customers to focus on building applications without worrying about provisioning and managing underlying Kubernetes clusters and associated infrastructure.
FIGURE 1. HPE GreenLake for ML Ops integrates various offerings in one prepackaged solution
By providing an enterprise-grade ML service offering that runs as a cloud in the customer’s data center, HPE GreenLake for ML Ops: • Allows data scientists to spend more time focused on data science by providing on-demand resources with hardware, infrastructure, and data science software as a service, and secure, self-serve provisioning and management via HPE GreenLake Central • Reduces operational risk by delivering as a fully managed service from control plane up to the infrastructure level and ML Ops-curated applications • Solves data gravity, data sovereignty, and compliance issues that are experienced with public cloud by keeping the data on premises and preinstalling open-source applications within the system • Reduces costs through flexibility to elastically scale up and down capacity usage dependent on needs and consumption-based pricing and billing • Offers a low-friction entry point with a trial experience option on prestaged hardware and software to test out the service before committing
FIGURE 2. A visual representation of the HPE GreenLake for ML Ops service solution and its components
