Oracle Explores Bring Your Own Hardware in Data Centers

Oracle may change how its cloud works by letting customers place their own hardware inside Oracle data centers.

This idea came up during an earnings call, where leaders said some large customers want more control over the machines they use.

Why customers are asking for this

Some companies already own servers and chips they trust.

They want to use that gear while still getting Oracle cloud services like databases and AI tools.

This could help customers manage cost and meet rules around data and control.

What it means for Oracle

Today, Oracle buys and installs all the hardware before selling cloud services.

If customers bring their own machines, Oracle would spend less upfront.

The way Oracle charges for cloud use could also change over time.

A new role for chip makers

Oracle said chip vendors may also place their own hardware in Oracle sites.

In this setup, the chip maker owns the gear and Oracle runs the cloud on top of it.

Other cloud firms have tried similar ideas before.

AI demand adds pressure

Oracle shared that demand for AI chips changes fast.

Customers often ask for GPUs, then change plans later.

Oracle shifts capacity between sites to keep up.

Cloud reach keeps growing

Oracle is also placing its database services inside Google Cloud data centers.

This gives customers more locations to run apps and store data.

If Oracle moves ahead, big cloud buyers may gain more choice and control over how cloud systems are built and paid for.

Red Hat shared its view on how hybrid cloud may look by 2026.

The post focuses on how companies are changing the way they run apps and data across data centers, public cloud, and edge sites.

Hybrid cloud becomes the default

Most companies no longer run systems in one place.

They mix on site systems, public cloud, and edge setups to meet cost, speed, and rule needs.

Red Hat sees hybrid cloud as the normal way forward, not a special case.

Open systems stay central

Red Hat stresses the use of open software and common tools.

This helps teams avoid lock in and move apps between places with less pain.

Linux and Kubernetes stay key building blocks for this setup.

AI changes how platforms are used

AI work puts new strain on systems.

It needs more power, faster links, and better control of where data sits.

Red Hat expects AI to push more work closer to data sources, not only into big clouds.

Security and rules drive choices

Rules around data use keep getting tighter.

Companies must know where data lives and who can reach it.

Red Hat sees security built into platforms rather than added later.

Teams need simpler tools

Hybrid setups add work for IT teams.

Red Hat points to a need for tools that are easier to run and easier to fix.

Automation helps teams manage large setups with fewer people.

The main message is clear: hybrid cloud will stay, grow, and shape how IT teams plan for the years ahead.

The Cloudcast explored whether AI agents can act like real experts, not just helpers.

The talk focused on how far AI is from doing full expert work across business fields.

What makes an expert

An expert knows rules, math, and standard practice.

That part can be taught using books and case studies.

Where AI falls short

Real experts gain skill from years of work.

They handle odd cases, gray choices, and tradeoffs.

Turning that kind of judgment into data is still very hard.

Big open questions

Who would train these agents and sell them.

How much an expert agent should cost.

How agents can learn without exposing private client data.

Where things stand now

AI works best today as a helper, not a replacement.

Coding tools are ahead, while other fields lag behind.

The key point is clear: expert level AI is coming slowly, and deep trust will take time.

Nutanix shared how its view on sovereign cloud is changing as more countries tighten data rules.

The focus is on control, location, and trust rather than just where servers sit.

What sovereign cloud means today

Many governments want data to stay inside their borders.

They also want clear rules on who can access that data.

This goes beyond simple hosting in a local data center.

Control matters more than labels

Nutanix says true control means customers manage systems themselves.

That includes software, updates, and access rights.

Just using a local cloud site is no longer enough.

Public cloud still plays a role

Sovereign setups do not always mean private cloud only.

Some public cloud tools can be used if rules are met.

The key is who runs the system and who can see the data.

Rules keep changing

Laws differ by country and keep shifting.

Companies must stay flexible to keep up.

This pushes demand for platforms that can run in many places.

What this means for IT leaders

Cloud choices now tie closely to law and risk.

Teams must plan for mixed setups that can change fast.

The main takeaway is clear: sovereign cloud is about control and trust, not just location.

NATO sees cloud sovereignty as a key factor for military strength. The Ukraine war highlighted the need to quickly connect an act on information, not just collect it.

Cloud technology must support secure data storage, intelligence sharing, and fast decision-making across all 32 member countries.

Sovereignty Dimensions
NATO defines sovereignty in three ways: control over data and its location, how systems are operated, and maintaining operations even if a provider is withdrawn.

Achieving full sovereignty may slow innovation, so NATO plans a mix of globally connected clouds and isolated, highly secure environments.

Collaboration and Innovation
American and European companies are partnering to create jurisdictionally isolated clouds that maintain speed and innovation.

NATO stresses that digital sovereignty can coexist with cooperation through technical safeguards and open standards.

Industry, academia, and allied nations are all crucial to developing AI-supported command systems and quantum-resistant security.

Urgency and Agility
NATO officials stress the need for rapid action, tech-savvy procurement, and collaboration with startups and non-traditional tech suppliers.

The UK recently pledged additional funds for drone and AI technology to support faster innovation in defense.

Securing digital sovereignty is about balancing national control with allied cooperation, while ensuring technology evolves faster than emerging threats.

In 2025, Azure’s network has become a critical driver for cloud innovation.

The Fairwater AI datacenter in Wisconsin highlights this, connecting hundreds of thousands of GPUs with 800 Gbps links, achieving performance far beyond the world’s fastest supercomputer.

Networking is no longer just support—it enables AI, global connectivity, and hybrid cloud operations.

Hybrid and Global Connectivity

Azure expanded ExpressRoute and VPN Gateway capacities, supporting massive AI workloads and hybrid setups.

Virtual WAN simplified global network management, including routing internet traffic through central security hubs.

Azure Route Server now handles up to 500 virtual connections, improving integration with customer equipment and third-party appliances.

Resilience and Security

ExpressRoute Resiliency Insights helps enterprises test and improve network reliability, while Azure DNS and forced tunneling features enhance security and compliance.

These tools ensure high availability and protection without heavy manual oversight.

AI-Powered Management

AI is embedded in network management through Azure Copilot, helping teams optimize, troubleshoot, and apply best practices efficiently.

Networks can become smarter, adapting and self-optimizing with AI support.

Looking Forward

Networking is now strategic, supporting real-time global supply chains, low-latency gaming, and international AI training.

AI-enabled networks promise self-healing, self-tuning, and more intelligent operations, making cloud infrastructure faster, safer, and easier to manage.

Azure Networking positions the network as a core enabler of cloud and AI innovation at scale.