In an effort to bring the cloud experience of generative AI to on-premises enterprise data centers, Hewlett Packard Enterprises and Nvidia have co-developed a new offering for the HPE GreenLake platform. The portfolio, called Nvidia AI Computing by HPE, will be sold through HPE or its partner resellers and includes a range of products and services set to be released in full this fall. This was announced during the keynote at HPE Discover 2024 by CEO Antonio Neri.
The HPE Private Cloud AI is a platform-as-a-service offering that leverages HPE hardware and software, Nvidia GPUs, AI tools, and models to provide a managed private cloud service through HPE’s GreenLake. This service aims to facilitate generative AI creation, allowing users to scale hardware based on demand while keeping data on premises and under their control.
The collaboration has also led to updates in other parts of GreenLake, such as the addition of a GenAI copilot in OpsRamp and a refresh of server hardware to support more powerful Nvidia GPUs.
While enterprises are still exploring GenAI capabilities through hyperscaler services, many will soon require solutions to manage GenAI at scale. HPE aims to address this need through their GreenLake offerings, which not only include hardware but also mature data management and query software, providing a comprehensive solution for customers.
The new HPE Private Cloud AI comes in various hardware configurations to support different AI workloads, powered by Nvidia GPUs, software platforms, and AI models. Liquid cooling installation options, as well as energy management and reporting software, are included in all configurations for improved efficiency.
Additionally, OpsRamp now features an operations copilot to assist IT ops teams in addressing operational challenges. This feature is seen as essential for IT ops automation tools but is still in its early stages of development.
As enterprises explore GenAI infrastructure, the cost can quickly add up. However, the potential benefits of controlling data in-house and eliminating the need for extensive reskilling or new hires might outweigh the associated costs in the long run.