Businesses and public sector organisations in Papua New Guinea (PNG), Australia and New Zealand (ANZ) are preparing for a future powered by artificial intelligence (AI), data analytics, and cloud-first applications. To keep up, they need IT infrastructure that’s fast, secure and ready to handle real-world complexity.
Hewlett Packard Enterprise (HPE) is helping organisations in the region meet this challenge. In the lead-up to 2026, HPE introduced major upgrades across its product range including servers, AI-ready data centres, hybrid cloud platforms, advanced storage and networking.
This article breaks down those innovations, explains what they mean for technology decision-makers across PNG and ANZ, and highlights how Stream Tech Knowledge PNG Ltd — an HPE Silver Partner and award nominee is helping local enterprises turn this technology into real business outcomes.
A New Generation of Servers: HPE ProLiant Compute Gen12
HPE’s flagship innovation for 2026 is the ProLiant Compute Gen12 family. Revealed in the middle of 2025, these servers signify a significant generational advancement, incorporating Intel Xeon 6 and AMD 5th-generation EPYC processors, direct liquid-cooling (DLC) options, and an AI-powered management framework.
Built‑in Security and AI‑Driven Management
Security is a priority for organisations across PNG and ANZ. HPE’s Gen12 servers incorporate the HPE iLO 7 Secure Enclave, offering quantum‑resistant cryptographic protection and meeting FIPS 140‑3 Level 3 standards. The servers are coupled with HPE Compute Ops Management (COM), an AI‑driven service that provides energy‑efficiency insights, predictive maintenance and a map‑based view of server health. According to HPE, the new platform uses AI to drive continuous optimisation and security across the lifecycle.
Availability Timeline and Models
Six of the eight Gen12 models are expected to launch in the first quarter of 2025: DL320, DL340, DL360, DL380, the AI‑focused DL380a and ML350. Larger models like the DL580 and Synergy 480 are expected to be released by mid-2025, targeting mission-critical and blade form factors. The ProLiant Compute Gen12 series offers optional direct liquid cooling for one- and two-socket systems, enhancing energy efficiency, which is an essential feature for data centers in hot regions such as Papua New Guinea.
Intel and AMD Editions
HPE presents the Gen12 servers equipped with either Intel Xeon 6 or AMD EPYC ‘Venice’ processors. The Intel models focus on quantum-safe security, integrated AI acceleration, and high-density computing capabilities. Meanwhile, the AMD-powered DL325 and DL345 Gen12 servers offer up to 6TB of memory and include HPE Morpheus VM Essentials for virtualization purposes. They also incorporate the same iLO 7 security and AI-driven management features as their Intel counterparts. These servers became available for order in mid-2025, with shipments commencing in July 2025.
DL380a Gen12: Designed for AI
One standout model is the DL380a Gen12, a 4U, two‑socket rack server built for AI workloads. It supports multiple double‑wide GPUs, high‑core‑count Xeon 6 CPUs (64–144 cores) and up to 4TB of RDIMM memory. This design is targeted at AI inference and training applications for enterprises. For organisations in PNG or remote areas of Australia, the ability to run AI workloads on‑site reduces reliance on internet connectivity and enables local data sovereignty.
Premier Solutions for Azure Local
Recognising the growing need for hybrid cloud, HPE introduced the ProLiant Compute DL380 Gen12 Server Premier Solutions for Azure Local, scheduled for availability in March 2026. This solution integrates Azure Stack HCI and provides a simple path to deploy on‑premises workloads with Azure services. It aligns well with organisations in PNG that want a cloud‑like experience while keeping data and compute within their own facilities.
Expanded GPU Support
HPE also announced expanded NVIDIA GPU options across its server portfolio: the RTX Pro 6000 Blackwell for the DL380 Gen12, RTX A1000 for the DL20 Gen11 and ML30 Gen11, and NVIDIA L4/L40 S options for the AMD‑powered DL325 and DL345 Gen12 servers. These options allow buyers to tailor compute for AI inference, graphics or general‑purpose acceleration.
Direct Liquid‑Cooled Supercomputing: HPE Cray Portfolio
In addition to enterprise servers, HPE unveiled its next‑generation Cray supercomputing portfolio, designed for AI and high‑performance computing (HPC). The announcement includes three direct liquid‑cooled blades—GX440n, GX350a and GX250—and a new Cray K3000 storage system.
Blade Specifications
GX440n: This top‑end blade contains four upcoming NVIDIA Vera CPUs and eight Rubin GPUs based on the NVIDIA Blackwell architecture. A single rack can house up to 192 GPUs, delivering enormous AI training capability. Cooled via direct liquid to enable high densities and energy efficiency.
GX350a: Targeting mixed workloads, the GX350a pairs AMD’s ‘Venice’ EPYC CPU with four AMD Instinct MI430X GPUs. Up to 112 GPUs can fit in a rack, offering strong floating‑point performance for HPC and AI.
GX250: This CPU‑only blade uses eight next‑generation AMD EPYC processors. It supports up to 160 CPU sockets per rack and suits simulation or analytics workloads that benefit from high memory bandwidth.
These blades work within a unified HPE Cray supercomputer architecture, controlled by a common software stack for compute, storage and networking management. They connect via the Slingshot 400 interconnect, providing high bandwidth and low latency for exascale workloads.
K3000 Storage with DAOS
HPE’s Cray K3000 storage system adopts the open‑source DAOS (Distributed Asynchronous Object Storage) platform. According to HPE, it offers high bandwidth and high I/O operations, making it ideal for data‑intensive AI and HPC workloads. The company emphasises that the system is factory‑built and modular, with energy‑efficient liquid cooling to reduce power consumption.
Energy Efficiency and Adoption
The new blades and storage are designed for energy-efficient supercomputing. HPE asserts that integrating direct liquid cooling with DAOS can enhance performance per watt by as much as 30-times over earlier solutions. These advancements have been embraced by prominent European HPC centers like the High-Performance Computing Center Stuttgart (HLRS) and Leibniz Supercomputing Centre (LRZ). For ANZ and PNG, these instances illustrate how HPE’s technology can support local research facilities or projects focused on weather and resource modeling.
AI Factory and GreenLake Innovations
AI Factory: Bringing Data and Compute Together
In late 2025 HPE introduced a portfolio of products described as the AI factory. These solutions support the full lifecycle of AI—from data ingest to model training and deployment. Key components and their availability include:
- HPE Data Fabric Software: A unified data plane that orchestrates data across hybrid cloud. It includes built‑in agentic AI governance and is available for customers on 31 October 2025.
- HPE ProLiant Compute XD685: A system with direct liquid cooling and NVIDIA Blackwell Ultra GPUs, targeted at AI training. It will ship in January 2026.
- NVIDIA GB300 NVL72 by HPE: A rack‑scale system combining NVIDIA GB200 chips, networking and software, engineered by HPE. Available to order and begins shipping in December 2025.
- DL380 Gen12 Premier Solutions for Azure Local: Mentioned earlier, this solution arrives in March 2026.
- Expanded NVIDIA GPU support across HPE’s server line: new Blackwell and Ada GPUs integrated into ProLiant and Apollo servers.
These components form a cohesive platform, allowing organisations to build on‑premises or hybrid AI infrastructures with HPE’s deep experience in supercomputing and direct liquid cooling.
GreenLake Intelligence and Agentic AI
At Stream Tech & HPE’s Tech Update Event in November 2025, HPE announced GreenLake Intelligence, an agentic AI framework designed to unify data and operations across networking, compute, storage and cloud platforms. GreenLake Intelligence collects trillions of metrics across HPE’s installed base and uses planning agents to coordinate data flows through a model context protocol (MCP). These agents sit between the infrastructure and AI models to provide context for operations like capacity planning, anomaly detection and cost optimisation. The framework aims to embed intelligence at every layer of the operational stack, enabling IT teams to act quickly and focus on innovation.
HPE Copilot for GreenLake, set to enter beta in late 2025, will provide a front‑end to this agentic platform, unifying hybrid cloud operations for customers. HPE emphasises that its agentic AI will appear across Aruba networking, OpsRamp (IT operations management), storage, cost‑operations, sustainability and even business services. For PNG and ANZ businesses, the promise is a single dashboard that delivers insights on performance, cost and sustainability across on‑prem, colocation and public cloud resources.
Aruba Networking and AIOps
HPE’s agentic AI is already integrated into Aruba Networking Central, the cloud‑managed platform for wired and wireless networks. According to HPE, Aruba Central will leverage anonymised telemetry data to provide root‑cause analysis and autonomous remedial actions. This capability helps IT teams in Australia and PNG quickly troubleshoot Wi‑Fi issues at remote sites, mines or campuses. In addition, AI‑powered modules can proactively detect security threats, unusual traffic patterns or hardware failures and recommend actions.
Sustainability and Energy‑Efficient Design
Environmental concerns are increasingly important for organisations in PNG and Australia, where electricity generation is often carbon‑intensive and expensive. HPE’s 2026 product roadmap emphasises sustainability:
- Direct Liquid Cooling (DLC) is available across ProLiant Gen12 servers, the XD685 platform, Cray blades and NVL72 systems. DLC allows servers to run at higher densities while using less power to cool them. For example, the Osaka data centre uses hybrid cooling to reduce environmental impact.
- Energy‑Efficient Supercomputing: HPE claims the new Cray platform can deliver 30× performance per watt improvements over previous architectures.
- GreenLake Sustainability Dashboards: Through GreenLake Intelligence and cost‑operations modules, customers can monitor CO₂ emissions and energy usage. This is valuable for achieving sustainability targets and reporting requirements.
Stream Tech Knowledge PNG, as a recipient of HPE’s Silver Partner Award and current nominee for the 2025 Hybrid IT Rising Star Partner of the Year, Stream Tech Knowledge PNG helps organisations in Papua New Guinea design data centres and server rooms that harness the benefits of Direct Liquid Cooling (DLC) while adapting to local climate conditions and power infrastructure.
Local Performance, Global Impact
With these innovations, HPE positions itself as an enabler for enterprises at the edge of the Pacific. In PNG and smaller Australian territories, connectivity and latency can be challenging. The new ProLiant Gen12 servers and AI factory platforms allow businesses to run advanced workloads locally, reducing dependency on long‑haul connectivity. For example, a PNG telecom operator could deploy the DL380a Gen12 with RTX Pro 6000 GPUs to handle real‑time video analytics or 5G core functions. Mining operations in Western Australia or PNG’s Highlands could use the XD685 and NVL72 systems to analyse geological data in near real time without sending sensitive information off‑site.
Hybrid Cloud and Data Sovereignty
The DL380 Gen12 Premier Solutions for Azure Local and the agentic AI capabilities of GreenLake give organisations a way to maintain control over data while accessing cloud functionality. This is crucial for sectors like finance and government that have strict data residency requirements. By combining on‑premises infrastructure with Azure or AWS, companies in New Zealand can adopt cloud innovation while ensuring compliance with local regulations.
Future‑Proof Investment
Investing in Gen12 servers or Cray supercomputing now will pay dividends into 2026 and beyond. The ability to upgrade GPUs (to Blackwell or MI430X), adopt next‑generation AMD or Intel CPUs, and integrate with agentic AI ensures that infrastructure remains relevant as AI evolves. The adoption by world‑leading HPC centres like HLRS and LRZ provides assurance of long‑term support and performance.
Skills and Support
HPE’s innovations are complex; organisations need partners with expertise in deployment, optimisation and ongoing management. Stream Tech Knowledge PNG Ltd is a certified HPE Solutions Partner and has engineers trained in ProLiant, Aruba, Cray and GreenLake technologies. Stream Tech also offers certified training in Microsoft, Cisco and cybersecurity, helping to build local capacity across PNG. With support offices in Port Moresby and Sydney, Stream Tech can provide on‑site assistance and 24/7 remote monitoring across the PNG–ANZ region.
Next Steps
HPE’s 2026 roadmap introduces powerful innovations across servers, supercomputing, AI factories, cloud platforms and sustainable design. For technology buyers in Papua New Guinea, Australia and New Zealand, these updates mean more choice and capability for modernising infrastructure and embracing AI. The ProLiant Compute Gen12 family delivers secure, AI‑optimised servers suitable for data centres and edge deployments. The new Cray blades and K3000 storage bring exascale‑class performance into reach, while the AI factory solutions and GreenLake Intelligence provide platforms for building and managing sophisticated AI workloads.
Stream Tech Knowledge PNG Ltd is positioned to help customers in PNG and the wider ANZ region navigate these options. As a certified HPE partner, Stream Tech can design, supply and support solutions tailored to local needs—whether you’re modernising a small data centre in Port Moresby, equipping a mine with on‑prem AI, or rolling out hybrid‑cloud services for a bank in Sydney. With the right combination of technology and expertise, the journey to an AI‑ready infrastructure in 2026 and beyond becomes achievable.