
Meta's Prometheus Supercluster
Meta plans to scale its Ohio AI cluster to 1 million GPUs.
ddNews Β· news about Hewlett Packard Enterprise (HPE) β products, AI,

Meta plans to scale its Ohio AI cluster to 1 million GPUs.

Internal documents reveal a $5bn order for Nvidia GB300 racks, but energy supply is failing to keep pace.

New capabilities for on-premises and air-gapped cloud environments to increase data control.

Analysis questions why Wall Street remains cautious despite HPE's surging earnings and growth.
quality(80)/p7i.vogel.de/wcms/7d/bf/7dbfd1f6386dae6bf8f3852551b84e89/0133876016v1.jpeg) β Analysis suggests enterprises are moving toward on-prem AI to reduce costs and external dependencies.

Funding focused on data centers and AI infrastructure through 2027.
The merger accelerates the rollout of high-speed 1.6T switches to meet AI data center demands.

A new diffusion LLM claiming to outperform GPT-6 Luna with higher speeds.

New vibration-based cooling for data centers and edge devices.
A new national HPC facility to drive AI research and disease diagnosis.
New system reaches 85.69 petaflops, enabling massive computational leaps.
New standard for optical circuit switching to improve AI workload economics.

New controls for governments to manage access and run AI without external connectivity.
Joint architecture unifies multi-agent orchestration and model routing for regulated workloads.
New secure workloads using CNCF confidential containers in partnership with Nvidia and AMD.
Potential second massive compute agreement for AI capacity.

Effort to secure valuation above $1.4 trillion ahead of a potential Anthropic IPO.
the stock saw a 3.18% drop following a rating downgrade.

An agreement reached at the Berlin investment summit to integrate AI into German industry.

New tools allow companies to measure the cost, ROI, and security of AI agents in a vendor-neutral way.
New partnership expands HPE distribution across Switzerland and Germany.

New AMD EPYC-based servers designed specifically for agentic AI, HPC, and high-performance analytics.
Nvidia releases a 120B parameter open-source model designed to power persistent, multimodal agents.

Westcon-Comstor becomes an official HPE networking distributor in Germany and Switzerland.
The new servers integrate 6th Gen AMD EPYC processors to boost performance.
HPE is deploying an advanced compute module to power the first high-performance computing (HPC) and AI workloads on the Moon via Astrolabβs FLIP rover.
New initiative designed to help enterprises move AI from experimental labs into production environments.
HPE relaunches the hyperconverged SimpliVity PC1000 with dual-hypervisor support in version 7.0.
Analysis of HPE's 400% growth and backlog versus market valuation.
Market shift toward alternative platforms as enterprises seek more flexibility for AI workloads and containers.
Market analysis suggests HPE stock may be undervalued by 28% following strong Q2 earnings.
HPE introduces a new server lineup specifically targeting next-generation AI workloads.
Greece inaugurates the β¬58.9 million Daedalus system, ranking 31st globally to support national AI and research.
Nvidia's CEO urges German manufacturers to integrate AI with physical heavy industry.

Four autonomous agents based on Gemini 4 successfully completed Minecraft without human intervention.
The AMD and HPE-powered supercomputer is delivering critical breakthroughs in tsunami prediction and drug discovery.

New York University deploys a new high-performance computing system to accelerate academic discovery.

Partnership with Miami HEAT to improve wireless reliability and fan digital experiences.
New APU featuring 128 GB of unified memory for high-end local AI workloads.
HPE is implementing a "Self-Driving Network" with Wi-Fi 7 for the Miami HEAT to enhance fan experience through autonomous connectivity.

Expansion of partnership to build agentic AI via the Gemini Frontier Capability Group.
HPE will deploy its Self-Driving Network and Wi-Fi 7 infrastructure for the 2027 Ryder Cup.
The Miami project highlights the strategic positioning of combining both Aruba and Juniper technology lines.
HPE and Miami HEAT partner to improve wireless reliability and connectivity at the arena.

Cognizant is dedicating 18,000 staff to build agentic AI on Google's Gemini platform.

Vultr updates its platform to improve security and execution for enterprise AI agents.
HPE shares are rebounding alongside Dell and Super Micro after a period of heavy sell-offs.

Modal reports that Nvidia's Arm-based Vera CPU doubles throughput for AI sandboxes compared to x86.

HPE is deploying an AI-native "Self-Driving Network" and Wi-Fi 7 for the Miami HEAT to enhance fan connectivity.

New support for CNCF Confidential Containers to secure Kubernetes AI workloads.

Deutsche Telekom, Orange, TelefΓ³nica, and Vodafone are partnering to boost European sovereignty against Starlink.
quality(80)/p7i.vogel.de/wcms/21/cd/21cd37fdcd7e641fb7c927555b2bc429/0134015374v1.jpeg) β A high-profile "Invest in Germany" summit focusing on industrial AI.

Amazon and AEP may deploy up to 960MW of X-energy small modular reactors for power.

A new open-source backend allows PyTorch code to run on AWS Trainium chips with minimal changes.

New tools launched to test AI agents for prompt injection and track operational costs.
HPE is rolling out next-generation Wi-Fi 7 infrastructure for the city.
quality(80)/p7i.vogel.de/wcms/0e/c6/0ec656ccdd52f8dbdb2630dd3e3e6640/0133878742v1.jpeg) β Techconsult study highlights the shift toward sovereign European platforms for AI workloads.
Goldman Sachs raised its HPE price target from $75.00 to $80.00.
The 85.69 petaflop system aims to make Greece a leader in European AI research.

The AI chip startup reached a $6.2bn valuation to develop its Galaxy servers.
Barclays lowered its HPE target to $77.00 following fiscal Q3 revenue of $12.21 billion.
HPE and Deliverance AI have validated a new architecture for deploying AI agents within private cloud environments.
HPE outlines key strategies for its networking business to drive long-term value.
HPE is offering the first year of virtualization services for free to attract customers.
Barclays lowered its HPE target to $77.00 following fiscal Q3 revenue of $12.21 billion.
quality(80)/p7i.vogel.de/wcms/e1/10/e1100c8deb1fd51b7ba492d4410f1edf/0133915175v1.jpeg) β Vertiv is moving toward delivering fully integrated data centers directly from the factory to reduce "time to token."
quality(80)/p7i.vogel.de/wcms/af/22/af22cce4f871fedc68a205a6d95f7ed2/0133578624v2.jpeg) β A new partnership unifies physical security infrastructure management into a centralized cloud IT model.

Case study shows enterprises moving AI to private clouds to control token costs, with HPE's smart routing saving up to 60%.
HPE identifies a growth opportunity in the refurbished market for Juniper Networks hardware amid the AI surge.

The first notebooks featuring Nvidia's Arm-Blackwell SoC have entered the market.