HPE lets loose VM Essentials to run on third-party platforms

HPE is continuing its GreenLake push with availability of its own virtualization product, plus disconnected operations for regulated environments, while touting an object storage platform intended to deliver greater performance and scale for modern workloads.

At the IT giant's Discover forum in Barcelona, the focus is once again on hybrid cloud and how HPE aims to meet its enterprise users' needs in such an environment.

Chief technology officer Fidelma Russo claimed the company has now assembled a comprehensive suite of tools to support end-to-end hybrid cloud operations, making it easier for IT teams to manage cloud-native and AI services alongside their traditional on-prem applications.

One recent acquisition was Morpheus Data and its cloud management capabilities, which HPE is now combining with its homegrown virtualization offering, disclosed at the company's Discover shindig in Las Vegas in June, into HPE VM Essentials, which is set to be available from December.

On top of that, while HPE virtualization was announced as an embedded technology within its private cloud platforms, HPE VM Essentials will also be available as a standalone version that can run atop third-party infrastructure as well as the company's own gear, indicating a potential appeal to disgruntled VMware customers.

"Now, regardless of where VM Essentials runs, whether it's private cloud, our hardware or third-party hardware, thanks to Morpheus, we can enable our customers to manage all of those VMs in a single management plane with Morpheus Data," said Hang Tan, COO of Hybrid Cloud.

This move was based on feedback following Discover Las Vegas, according to Tan, who claimed that beta testers pronounced it ready to deploy in production environments right away, and expressed approval of the per-socket licensing HPE is proposing. Not that the company has disclosed licensing costs yet.

VM Essentials can also manage existing hypervisor clusters, and features image conversion capabilities, according to HPE. Tan also said that data protection firms Cohesity and Commvault will support the platform, as well as it integrating with HPE's own tools such as Zerto and OpsRamp.

One notable development the company has made for customers with strict security and regulatory compliance requirements is an option for the GreenLake platform to operate in a disconnected mode with no internet connection.

This includes HPE Private Cloud Enterprise and Alletra Storage MP for disconnected datacenter environments, but still delivering a cloud-like operating experience.

Many such customers also have sovereign requirements with regard to their data, according to HPE's GM for Private Cloud and Flex Solutions, Cheri Williams, who said that select partners in specific regions would be able to build a sovereign cloud using Private Cloud Enterprise Disconnected as a managed solution.

No HPE event would be complete without a mention of AI, and the firm is expanding on the turnkey private cloud for AI (PCAI) offering in partnership with Nvidia it unveiled at Las Vegas in June. Now Deloitte has been added to the mix, Williams said.

"Deloitte has a number of AI offerings that they're working on, one of which is C-Suite for CFOs," she said, citing a collection of AI use cases, including financial statement analysis, along with another two that have now been enabled on PCAI and available to customers.

• HPE nabs long-time ally Morpheus Data
• HPE GreenLake sales on the rise as branding tweaks include more lines
• HPE, Nvidia team up to offer 'turnkey' private cloud racks for keeping AI loads local
• HPE Q2 sales rise was brought to you by the letters A and I

Also unveiled at Discover is the Alletra Storage MP X10000, an unstructured data platform for modern workloads that HPE grandly proclaims as unifying its storage portfolio by adding object capabilities to the existing block and file data services already supported.

The MP X10000 has been designed for exabyte scale and simplifies management across hybrid environments thanks to the firm's multi-protocol (MP) disaggregated architecture, according to SVP & GM for HPE Storage Jim O'Dorisio.

"With zero over provisioning and a purpose built set of data services on the Alletra Storage MP architecture, we feel like we have differentiated capabilities that enable customers to take full advantage of block, file and object across a single architecture," O'Dorisio said, claiming that it offers a 40 percent total cost of ownership (TCO) advantage over monolithic architectures, plus a reduction in energy consumption.

With fast all-flash object storage, HPE is targeting two primary use cases – data lakes and data protection. The MP X10000 is claimed to provide ample bandwidth for rapid backup and recovery of mission-critical applications and up to 20x data reduction for better integration with third-party backup solutions. It has been certified by partners including Cohesity, Commvault, and Veeam to work with their services.

O'Dorisio also claimed HPE is working with Nvidia to enable a direct memory access pathway between GPU memory, system memory, and the MP X10000, to improve bandwidth and reduce latency for AI workloads using an interface compatible with the Amazon S3 storage protocol.

"This will be the first of many new and exciting features that we will bring the market over the course of 2025," he said.

Speaking of GreenLake itself, Russo said the platform had evolved greatly since it was first announced five or six years ago.

"What first started as a financial structure to really deliver consumption opportunities to on-prem customers, to have the same kind of financial structure and pay-as-you-go as the public cloud, has really transformed into a platform based cloud that embraces infrastructure, not just in your datacenter, but in your colo, at the edge, and as appropriate, on the public cloud," she explained.

"So we've evolved from a financial structure to a platform structure, and you can see that now we are starting to really bring true innovation to our customers as they're grappling with their traditional workloads, and also with their AI workloads going forward." ®