HPE Enhances Hybrid Cloud with AI-Powered GreenLake

Hewlett-Packard Japan explained its integrated hybrid cloud platform strategy. HPE Executive Officer and General Manager of Pre-Sales, Nakamura Fumihiko, said, “By integrating agent-based AIOps into almost all infrastructure layers, HPE can transform GreenLake Cloud into a hybrid cloud that utilizes agent-based AI. Going forward, we will implement agent-based AI in each layer of GreenLake, including infrastructure, data platform, runtime software, and operations.”

Integrated Hybrid Cloud Platform StrategyOn the 29th, HPE LLC explained its integrated hybrid cloud platform strategy. HPE Executive Officer and General Manager of Presales, Fumihiko Nakamura , stated, “By integrating agent-based AIOps into almost all infrastructure layers, HPE can transform the GreenLake cloud into a hybrid cloud that utilizes agent-based AI. Going forward, we will implement agent-based AI in each GreenLake layer, including infrastructure, data platform, runtime software, and operations.”

This framework is called “HPE GreenLake Intelligence.” Nakamura, Executive Officer of Hewlett-Packard Japan, stated, “GreenLake Intelligence will transform hybrid cloud operations using agent-based AIOps.” He used the metaphor, “If GreenLake is a car, then GreenLake Intelligence is like a car equipped with cutting-edge autonomous driving capabilities.” AI agents in various layers, such as infrastructure, operations, and FinOps, act as assistants to help companies achieve their operational policies and service levels. They can provide insights and quickly resolve problems. “When provisioning infrastructure, agents confer with each other and present configuration proposals based on performance, cost, and other factors. Administrators can select the optimal proposal and begin using it. During operation, the system detects and diagnoses performance issues and presents solutions. This improves productivity, reduces risk, speeds repairs, and minimizes failures throughout daily operations and the entire operational lifecycle.” He also noted, “By using MCP, you can collaborate with specialized inference agents provided by GreenLake Intelligence.

こちらもお読みください： RIKEN teams with Fujitsu & NVIDIA on Fugaku NEXT system

In the future, we will release agents for servers, storage, networks, OS, observability, FinOps, and more. These agents will work together to achieve results in a full-stack environment.” He also mentioned that by implementing GreenLake Copilot in the GreenLake portal, users can manage and operate increasingly complex hybrid cloud environments using natural language. The first implementation of GreenLake Intelligence is planned for HPE Aruba Networking Central, where it will analyze data collected from approximately 6 million network devices and over 3 billion client endpoints. It will utilize the rich context provided by HPE’s Agent Mesh to perform autonomous inference, enabling accurate cause identification and repair. “The first product will be GreenLake Intelligence for HPE Aruba Networking Central. More than 15 specialized AI agents, categorized as real-time and knowledge-based, will work together to continuously optimize the system and notify administrators. It is scheduled to be available from the third quarter of 2025.” The second, “HPE CloudOps Software,” is a new software suite that enhances existing hybrid cloud operations. It consists of “HPE Morpheus,” which provides integrated multi-cloud management; “HPE OpsRamp,” which enables full-stack multi-vendor AIOps; and “HPE Zerto,” which protects data from edge to cloud and detects ransomware in real time. “By combining these products, which were previously offered individually, we can benefit from the enhancement of operations in hybrid cloud environments. While each product is from a company acquired by HPE, they share common elements of being multi-vendor and multi-cloud, allowing them to be used together. This will increase flexibility through solutions and accelerate value creation.” Furthermore, HPE announced the integration of HPE Morpheus VM Essentials with HPE Private Cloud. He also announced the implementation of microsegmentation with the CX10000 and integration with Veeam, Commvault, and Cohesity. “We’re also launching services to modernize and migrate existing virtualized environments. VM Essentials now supports new hardware, including third-party products like Dell PowerEdge and NetApp AFF. We’ll continue to upgrade the platform,” he said. The third offering, the “AI Factory Solution,” is positioned as a solution for addressing increased AI workloads. “It addresses how to think about and operate AI across the entire company and across departments. It covers everything from the pilot phase to profit generation and results, guiding successful implementation and utilization of AI and enabling the ultimate hybrid workload operation.”

HPE Private Cloud AI, developed in collaboration with NVIDIA, offers the Turnkey AI factory for enterprises; the AI ​​factory at scale for large-scale AI deployments aimed at service providers and model developers; and the Sovereign AI factory for handling highly confidential data from government and public institutions, finance, pharmaceuticals, and other sectors. The company also demonstrated the ability to use HPE CloudOps Software as a common control plane across the entire portfolio. It also announced that HPE OpsRamp has been certified as an observability solution for NVIDIA Enterprise AI Factory. “It can provide full-stack observability in multi-vendor environments, not just HPE components,” the company said. Furthermore, as part of the Next-Generation HPE Private Cloud AI, it protects sensitive data with an air-gapped configuration that is not connected to the internet. Multi-tenancy also supports group collaboration, GPU quota management, and role-based access control in the enterprise. Furthermore, HPE’s patent-pending federated architecture allows for incremental investments and ensures continuous protection of AI investments. Additionally, HPE Private Cloud AI offers 75 validated AI use cases and announced collaborations with 26 new AI ISV solution partners, including Deloitte and Accenture .

At HPE Discover 2025, held in Las Vegas in June 2025, the company announced the expansion of its portfolio, including “HPE GreenLake Intelligence,” which revolutionizes hybrid cloud operations with agent-based AIOps; “HPE CloudOps Software,” which improves flexibility and accelerates value creation through software solutions; and “AI Factory Solutions,” which provide solutions across cloud, core, and borders to meet all AI goals. These expansions demonstrate the company’s intention to accelerate its integrated hybrid cloud platform strategy. “Most companies now operate hybrid clouds that combine on-premises and public clouds, and are entering a phase of operational sophistication. Furthermore, demand for AI is increasing. HPE calls this the ultimate hybrid workload,” HPE stated. “This brings with it new challenges, such as the integration of distributed data, secure data management, building infrastructure optimized for AI processing, and power consumption issues. HPE believes it is necessary to equip its platform to accommodate new AI-driven workloads. This is the fundamental strategy of HPE’s integrated hybrid cloud platform.” The core of the company’s integrated hybrid cloud platform strategy is the agent-based AI framework “HPE GreenLake Intelligence.” Nakamura, Executive Officer of Hewlett-Packard Japan, stated, “GreenLake Intelligence will revolutionize hybrid cloud operations using agent-based AIOps.” He likened it to a car, saying, “If GreenLake is a car, then GreenLake Intelligence is like a car equipped with cutting-edge autonomous driving capabilities.” AI agents in various layers—infrastructure, operations, and FinOps—serve as assistants, helping companies achieve their operational policies and service levels. They can provide insights and quickly resolve problems. “Even during infrastructure provisioning, agents confer with each other and present configuration proposals based on performance and cost, allowing administrators to select the optimal proposal and begin using it. Furthermore, during operation, performance issues are detected, diagnosed, and solutions are presented. This enables improved productivity, reduced risk, faster repairs, and minimized failures throughout daily operations and the entire operational lifecycle.” He also said, “By using MCP, you can collaborate with the specialized inference agents provided by GreenLake Intelligence. In the future, we will release agents for servers, storage, networks, OSs, observability, FinOps, and more, and these agents will work together to achieve results in a full-stack environment.” He also mentioned that by implementing GreenLake Copilot in the GreenLake portal, it will be possible to manage and operate increasingly complex hybrid cloud environments using natural language.

The first implementation of GreenLake Intelligence is planned for HPE Aruba Networking Central, where it will analyze data collected from approximately 6 million network devices and over 3 billion client endpoints. It will accurately identify the cause and perform repairs through autonomous inference, leveraging the rich context provided by HPE’s agent mesh. “We will initially commercialize it as GreenLake Intelligence for HPE Aruba Networking Central. More than 15 specialized AI agents, categorized as real-time and knowledge-based, will work together to continuously optimize the system and notify administrators. It is scheduled to be available from the third quarter of 2025.” The second, “HPE CloudOps Software,” is a new software suite designed to enhance existing hybrid cloud operations. It consists of “HPE Morpheus,” which provides integrated multi-cloud management; “HPE OpsRamp,” which enables full-stack, multi-vendor AIOps; and “HPE Zerto,” which protects data from edge to cloud and detects ransomware in real time. “By combining these products, which were previously offered individually, we can enhance operations in hybrid cloud environments. While each product is from a company acquired by HPE, they share common multi-vendor and multi-cloud elements and can be used together. This will increase flexibility and accelerate value creation through solutions,” the company said. Furthermore, the company announced the integration of HPE Morpheus VM Essentials with HPE Private Cloud. It also announced the implementation of microsegmentation with the CX10000 and integration with Veeam, Commvault, and Cohesity. “We will also launch a service to modernize and migrate existing virtualized environments. Additionally, third-party products such as Dell PowerEdge and NetApp AFF have been added to the list of hardware supported by VM Essentials. We will continue to provide incremental upgrades,” he said. The third offering, the “AI Factory Solution,” is positioned as a solution for expanding AI workloads. “It will address how to think about and operate AI when applying it to the entire company or department. It covers everything from the pilot phase to profit generation and results, leading to successful AI implementation and utilization and enabling the ultimate hybrid workload operation.” HPE Private Cloud AI, developed in collaboration with NVIDIA, offers the Turnkey AI factory for enterprises, the AI ​​factory at scale for large-scale AI targeted at service providers and model developers, and the Sovereign AI factory for handling highly confidential data from government and public institutions, finance, pharmaceuticals, and other sectors. HPE also highlighted the ability to use HPE CloudOps Software as a common control plane across the entire portfolio. The company also announced that HPE OpsRamp has been certified as an observability solution for NVIDIA Enterprise AI Factory. “It can provide full-stack observability across multi-vendor environments, not just HPE components,” the company said. Furthermore, the company announced the Next-Generation HPE Private Cloud.For AI, the air-gapped, internet-disconnected configuration protects sensitive data. Multitenancy supports enterprise group collaboration, GPU quota management, and role-based access control. HPE’s patent-pending federated architecture enables incremental investments and continuously protects AI investments.

HPE Private Cloud AI also offers 75 validated AI use cases and announced collaborations with 26 new AI ISV solution partners. AI agents for financial services provided by Deloitte and Accenture are also available. HPE also announced collaborations with SCSK and CTC in Japan regarding HPE Private Cloud AI. “We will provide trial environments in Japan and strive to expand the ecosystem,” the company said. The company also announced Release 2 of its HPE Alletra MP Storage X10000, an agent-driven storage system specialized for AI workloads. It incorporates Data Intelligence functionality, enabling it to add contextual information to data and communicate with external LLMs in natural language. The company also announced native implementation of MCP. He added, “HPE has built the top three supercomputers on the Top 500 list, and is highly rated for its energy efficiency. This track record and cooling technology know-how can be utilized when implementing large-scale AI environments.” He also stated that HPE Services offers Unified Cloud Management, which allows customers to select and implement the optimal hybrid cloud operation method. “You can define and implement the optimal operation method for eight operational areas: PlatformOps, AIOps, FinOps, SSAMOps, SecOps, GreenOps, NetOps, and DataOps. Utilizing a unique maturity model, it is possible to promote optimal application.”HPE also offers AI agents for financial services provided by HPE. HPE also announced collaborations with SCSK and CTC in Japan on ヒューレットパッカード Private Cloud AI. “We will prepare trial environments in Japan and work to expand the ecosystem,” the company said.

HPE also announced Release 2 of its agent-driven storage system, the HPE Alletra MP Storage X10000, specifically designed for AI workloads. It incorporates Data Intelligence functionality, allowing it to add contextual information to data and interact with external LLMs in natural language. It also announced native implementation of MCP. Furthermore, HPE stated, “HPE has built the top three supercomputers on the Top 500 list, earning high praise for their energy efficiency. This track record and cooling technology expertise can be leveraged in the deployment of large-scale AI environments.” HPE Services Unified Cloud Management is also available, allowing customers to select and implement the optimal hybrid cloud operation method. “It is possible to define and aim for implementation of optimal operational methods for eight areas that require operational consideration: PlatformOps, AIOps, FinOps, SSAMOps, SecOps, GreenOps, NetOps, and DataOps. By using a unique maturity model, it will be possible to promote optimal application.”

ソース ヤフー