🧠 Editorial

While outside the asphalt is slowly turning into lava, the tech industry is once again discussing the truly important questions: How sovereign is Europe’s cloud? Do we need our own hyperscalers? And why is everything getting more expensive – including computing power?

This issue certainly has enough material for the next coffee break. It covers cloud dependencies, German alternatives, rising prices, and the realization that “we’ll deal with it later” is rarely a reliable infrastructure strategy.

Also on board: Sven Hennessen, who as a guest author talks about the DWX Developer World. Anyone wanting to know what developers, architects, and platform teams are really talking about should definitely not skip this article.

So, quickly adjust the fan and let’s get started.

📰Tech-News:

A “German Hyperscaler” Won’t Solve Your Problem

The demand for German hyperscalers initially sounds logical. After all, those who want digital sovereignty do not want to be dependent on AWS, Microsoft, or Google. But the real problem often lies elsewhere.

Most companies do not operate a streaming platform with hundreds of millions of users. They are not Netflix. They do not process data volumes that require a global infrastructure with hundreds of thousands of servers. Yet, purchases are often made as if that were the case.

When companies are asked about their requirements, surprisingly often no requirements are mentioned, but product names. Microsoft Azure. Microsoft 365. Copilot. The entire Microsoft world.

That is not a clearly defined need. That is the adoption of a vendor portfolio.

At this point, it is often argued that European alternatives would not suffice. But what does that actually mean?

Is a technical function really missing? Or is it just the identical interface, the identical product name, and the identical marketing promise that are missing?

Those who want digital sovereignty must first honestly ask themselves what requirements actually exist. Because there is a world of difference between “We need exactly product X” and “We need a secure collaboration platform.”

The crucial question is not which provider has the longest feature list. The crucial question is: What is actually needed? How many users are there? Which applications are business-critical? What availability is really required? Which data needs special protection? Only when these questions are answered can it be assessed which infrastructure is necessary.

The reality is much less spectacular than many marketing slides suggest. A large part of corporate applications can be easily operated on European infrastructure. With open standards. With Open Source. With significantly more control over data, costs, and dependencies.

However, this requires being willing to critically question one’s own needs instead of reflexively ordering from a hyperscaler’s catalog. Yes, Germany needs powerful cloud providers. Because Europe needs more digital sovereignty.

But a German hyperscaler does not solve a problem if companies continue to purchase technologies they never needed. Sovereignty does not begin with the provider. It begins with an honest analysis of one’s own requirements.

🔗https://www.heise.de/news/Unternehmen-wuenschen-sich-deutsche-Hyperscaler-und-buchen-US-Cloud-11335451.html?utm_term=Autofeed&utm_medium=Social&utm_source=LinkedIn#Echobox=1781692043

Hetzner’s Price Increase Shows How Dependent Europe Really Is

The discussion about the new Hetzner prices has so far been surprisingly superficial. Understandably, many initially look at the numbers. When individual cloud products suddenly cost twice or even three times as much as before, it quickly gives the impression that a provider has turned the price screw because the market currently allows it.

I believe this is the wrong interpretation.

Because the crucial question is not why Hetzner increased its prices. The crucial question is rather why a company that has been known for an exceptionally good price-performance ratio for many years and whose success was largely based on exactly this promise, is forced to take such a step at all.

The answer begins with the hardware.

Today, those who sell cloud servers ultimately sell computing power, storage, and network resources. However, behind these abstract products are physical components whose procurement has not only become more expensive in recent years but also strategically more complicated. Modern processors, high-performance storage, SSDs, or network technology are created in global supply chains that are concentrated on a few manufacturers, a few regions, and sometimes even on individual production sites.

The current AI boom further exacerbates this situation. While classic hosting providers used to mainly compete with other hosting providers for components, today they compete with companies that invest billions in AI infrastructure and are willing to pay almost any price for available capacities. The consequence is foreseeable: Those who build or operate servers must accept higher procurement costs.

In this context, the price increase at Hetzner appears less as a corporate decision than as a symptom of a development that began far outside the company’s sphere of influence.

However, the story does not end there.

Because Hetzner faces a challenge that, in my view, is much more interesting than the actual price adjustment.

For many years, the company could rely on the combination of solid technology, a European location, and very attractive prices to create a convincing overall package. Especially compared to the American hyperscalers, the argument was often not that Hetzner offered more features, better support, or a larger ecosystem. The strongest argument was usually that you got an extraordinary amount of infrastructure for your money.

If prices now move significantly upwards, the basis of this comparison inevitably changes.

The smaller the price gap becomes, the more other criteria come to the fore. Companies then begin to evaluate more precisely which additional services are available, how flexibly platforms can grow, what integrations exist, and how high the operational effort is in daily operations. In other words: A provider that has long convinced primarily through an excellent price-performance ratio must now increasingly convince through the quality of its overall offering.

This is not a reproach to Hetzner. Rather the opposite.

Because the real irony is that one of the most successful European infrastructure providers is currently feeling the effects of those dependencies that have been discussed in Europe for years. We regularly talk about digital sovereignty, technological independence, and strategic resilience. At the latest, when rising hardware prices directly reach European providers and their customers, it becomes visible that these debates are not political fringe topics but economic reality.

For me, the Hetzner price increase is therefore primarily one thing: a reminder of how expensive technological dependency becomes as soon as theoretical risks turn into concrete bills.

🔗https://docs.hetzner.com/de/general/infrastructure-and-availability/price-adjustment

FISA 702 Expires. Relief for US Clouds? Not Really.

A central law of American foreign surveillance, the infamous Section 702 of FISA, has expired. Anyone who now believes that this ends the privacy and sovereignty debate around AWS, Microsoft Azure, or Google Cloud should take a closer look.

Because the real insight of the current discussion is another: The problem was never solely the law itself.

Leading US politicians are already signaling that the surveillance practice should continue as unchanged as possible. At the same time, the history of the past decades shows that surveillance powers in the US have been regularly adjusted, expanded, or replaced by new legal bases.

For European companies, this means above all one thing: The assessment of risks should not be based on individual laws but on the fundamental question of which jurisdiction a provider is subject to and what data is processed there.

US hyperscalers remain technically leading and are the right choice for many use cases. However, sensitive data, strategic company knowledge, research data, or information with special protection needs should continue to be subjected to a very careful risk assessment.

The expiration of FISA 702 does not automatically make American cloud providers problem-free.

Rather, it reminds us why digital sovereignty and data sovereignty remain important topics in Europe.

🔗https://www.heise.de/news/US-Spionage-Rechtsgrundlage-laeuft-aus-Spionage-laeuft-weiter-11329582.html?utm_term=Autofeed&utm_medium=Social&utm_source=LinkedIn#Echobox=1781236300

🗣️LinkedIn Post of the Week

A noteworthy post by Dominik Rapacki, CEO of meetergo, addresses a question that is surprisingly rarely asked in many discussions about European AI companies: Who actually owns these companies?

Using examples like Langdock, Black Forest Labs, and KugelAudio, he shows that many AI startups perceived as “European” are indeed developed in Germany or Europe, but their ownership structure is often organized via US holdings in Delaware. The background is usually the same: International investors and US venture capital funds require such structures to be able to invest.

The post becomes particularly interesting where Dominik draws the connection between ownership structure, CLOUD Act, and digital sovereignty. Because the question of whether data is stored in Europe is not necessarily identical to the question of which law a company ultimately falls under.

An important and nuanced perspective on the current debate around European AI champions, digital sovereignty, and what “Made in Europe” actually means.

🔗 https://www.linkedin.com/posts/dominik-rapacki_digitalesouveraeunitaeut-cloudact-ki-share-7474700472889769984-mAPB/?utm_source=share&utm_medium=member_desktop&rcm=ACoAADCSWyQBU4m7hUbXDJqk27ftrkLIYOZzONU