HPE Highlights Intelligent Edge and the AI Cloud

At its HPE Discover 2023 event this week in Las Vegas, Hewlett Packard Enterprise (NYSE: HPE) rolled out multiple GreenLake services and a blockbuster generative AI solution.

“The three key elements of our strategy are the edge, hybrid cloud, and artificial intelligence,” said HPE CEO Antonio Neri during his keynote address Tuesday. “And each of these are connected.”

Neri started, appropriately enough, at the edge. “The intelligent edge is where we live and work, and the global pandemic has transformed a legacy enterprise business into a distributed, digital enterprise business,” he said. That involves mobile networking, and Neri cited HPE’s Aruba division as a key tool in making the network an enabling infrastructure for security and connectivity to its GreenLake hybrid cloud services. Neri also mentioned HPE’s acquisition of Silver Peak in 2020 as part and parcel of the edge strategy. For both product lines, he said, HPE has offered AIOps (artificial intelligence for IT operations) and will continue to do so via the addition of upcoming natural language models.

Neri also promised that private 5G would be combined with WiFi in solutions based on HPE’s recent acquisition of Athonet, an 18-year-old company that makes core software for private 4G and 5G networks.

HPE CEO Antonio Neri. Source: HPE

Enterprise Workloads Remain Mostly In House

Despite a surge toward public cloud adoption, CEO Neri said that 79% of enterprise workloads remain on premises. Hence the need for a variety of private and hybrid cloud solutions. Leading the pack was a series of enhanced offerings with public cloud leader Amazon Web Services (AWS), as listed below:

• Generally available support for AWS Elastic Kubernetes Service (EKS) Anywhere on HPE GreenLake for Private Cloud Enterprise.
• Support in HPE GreenLake for Backup and Recovery for AWS EKS Anywhere and AWS Relational Database Service (RDS), along with support for Amazon Elastic Compute Cloud (EC2) instances and Amazon Elastic Block Store (Amazon EBS) volumes.
• Availability of the HPE NonStop Development Environment as an Amazon Machine Image (AMI) in the AWS Marketplace.
• Availability of HPE Fraud Risk Management as a service in the AWS Marketplace. This is a solution addressing fraud threats associated with 5G and Internet of things (IoT) environments, as well as a host of fraud types encountered by telcos.

HPE also announced that its HPE GreenLake for Private Cloud Enterprise and a new HPE GreenLake for Private Cloud Business Edition would be pre-provisioned in Equinix (Nasdaq: EQIX) IBX datacenters in seven metro areas to start – Washington, D.C., Silicon Valley, Toronto, London, Frankfurt, Singapore, and Sydney.

HPE GreenLake for Private Cloud Enterprise allows enterprise customers to configure a “self-service” assembly of choices among bare metal, virtual machine, and container workloads in a “pay as you go consumption model.” The new HPE GreenLake for Private Cloud Business Edition enables virtual machine functionality for “spinning up” configurations faster – a capability cited as important for integrators.

AI at the Apex

CEO Neri saved the hottest news for last – HPE GreenLake for Large Language Models (LLMs). “HPE is entering the AI public cloud market,” he declared, saying the announcement was among the most important the company ever made. “HPE GreenLake for Large Language Models will enable any enterprise to privately train, tune, and deploy large-scale AI through an on-demand, multi-tenant supercomputing cloud service.”

Initially, the service will be offered from a high-density colocation facility in Quebec owned by QScale, which specializes in environmentally sustainable computing centers. According to CEO Neri, the setup will be “nearly 100%” powered by clean energy. Availability of the LLM will start in North America this year and extend to Europe early in 2024, HPE said.

Neri stressed that the new service can be used either to build, train, and publish LLMs for use by other companies or to pre-train LLMs for internal use cases. In the future, HPE will publish specialized LLMs of its own tailored to specific enterprise apps.

The new service takes advantage of HPE’s high performance computing (HPC) and supercomputing expertise and infrastructure, enhanced by its acquisitions of Cray in 2019, Determined AI in 2021, and Pachyderm in January 2023.

Out of the gate (there will be other partners), HPE GreenLake for LLMs offers access to Luminous, an LLM from Aleph Alpha, a German startup whose CEO and founder Jonas Andrulis appeared on stage with Neri during the announcement. Luminous is capable of training models in multiple languages, including English, French, German, Italian, and Spanish. Andrulis stressed that it’s intended for sophisticated enterprise applications.

“What we are trying to do is we’re trying to bring this capability, these powers, into complex and critical environments, like finance, legal, health, government, security,” Andrulis said. “And for these environments you need more than just a chatbot that burps out an answer.”

Jonas Andrulis, founder and CEO, Aleph Alpha. Source: HPE

All told, HPE had a lot of news this week, all demonstrating its talent for acquisition and partnerships, its comprehensive enterprise focus, and its clever GreenLake packaging. The addition of an enterprise-grade LLM service ups the ante for competition from some of its ecosystem partners, such as AWS, Red Hat, and VMware, along with other public cloud providers, including Microsoft Azure and Google Cloud. How HPE navigates its way will determine its future.