GaiaX - a federal rocket
The federal rocket of Altmeier.
What is a federal cloud?
In simple terms, a cloud is a pile of IT services that are ordered, delivered and billed automatically. In other words, it's like Spotify: go to the website, enter your credit card, select the service, and get started. In a similar way, you can order storage, network and computing power from AWS. However, you wouldn´t be able to do this with an old-style data center or IT provider.
A federal cloud is when the customer can order IT services from several providers on the same portal using the same principle: The storage at T-Systems, the Kubernetes cluster at ATOS and the AI service at Bosch. In this way, many small clouds, each with a few services, become one large federal cloud. What's the advantage? Marco-Alexander Breit, the digitization expert of Minister of Economics Altmeier, puts it this way in the Handelsblatt:
"Individual lakes are to become a lakeland"
These many small lakes already exist, each one smaller than Google, Azure or AWS. But together it could become a huge lakeland.
What is the catch?
The idea sounds smart, but there might be a catch or two:
Customers cannot distribute many of their applications (e.g. SAP) to many small lakes at will. In order to compete with the central clouds, the respective sub-cloud provider must therefore be sufficiently attractive. And that means: many services that scale sufficiently and are not significantly more expensive than the competition. Translated to the sea metaphor, this means: A container ship will not be happy in Käbelicksee near Dalmsdorf, even if the overall lake land is 10 times larger than the Bodensee.
Entire IT landscapes, on the other hand, can certainly be distributed among many providers, but then other problems arise. Although all (large) clouds offer the same core, the details are very different. AWS specialists can bring applications to AWS, Google specialists to Google. So if the federal cloud is based on factually different ecosystems, companies have to train and utilize separate teams for each. And the sea metaphor? Each lake in the lake land has its own maps, its own driver's licenses, its own pilots.
If a CIO nevertheless distributes his IT landscape over different technical clouds, the applications will sometimes want to talk to each other. This requires secure connections between each location. This means setup effort, operating effort and usage-based costs. The lake metaphor: transporting containers from Priesterbäckersee to Hofsee is something completely different than delivering containers from Zenssee to Leppinsee.
What will it fail?
Making a decentralized cloud interoperable, implement billing and orchestrate a diver service catalogue will be a small moon landing in itself. But even if the GaiaX team succeeds in this, the federal structure will cause the customer significantly more effort than he would have to put into each of the three hyperscalers. The increased total cost of ownership is thus only countered by the vague promise of data sovereignty. My forecast on this from everyday company life: this alone will not be enough.
Why keep at it anyway?
The cloud market is gigantic - and will grow even more. The cloud is the digital infrastructure of the digital world of the future. Just as the Saudis earn money with every kilometer driven by a car, so the cloud providers earn money with every click a user makes. And unlike oil, it is not God-given whether we find it in Europe or not. So, dear GaiaX team, stay tuned, even if maybe one or two pivots will be necessary.
This article was supported by Andreas Tamm and Roland Frank