NVIDIA takes GTC to Berlin and bet on AI agency and European data sovereignty

NVIDIA held the GTC in Berlin and used the European stage to reinforce three fronts at the same time: scale AI computing, AI and data sovereignty. The movement expands the company's presence beyond the United States in full global race for processing capacity. According to official source of the announcement, the event combines infrastructure commitments in France with an ecosystem of partners that develops AI solutions on the continent.

What did NVIDIA GTC Berlin bring again?

The GTC is the company's technical conference, historically held in San Jose, California. Taking the format to the German capital means replicating in Europe the logic that supports the matrix event: platform announcements, technical demonstrations and approximation with the local ecosystem of developers, companies and infrastructure providers.

The timing is not casual. Berlin occurs after a sequence of global NVIDIA announcements, including the presentation of next-generation AI chips and robotic platforms on GTC 2026. The European edition acts as a regional development of this strategy, adapted to the demands of those buying computing on the continent.

Why is European data sovereignty a priority?

Sovereignty, in this context, means processing data and training models under European jurisdiction, with infrastructure installed on the continent itself. Governments and companies in the region began to treat computational capacity as a strategic asset, and not only as an IT item.

NVIDIA is a provider of the computing layer for these projects. The announcement cites infrastructure commitments in France, a theme already indicated in the billionaire plan signed with the French government. The logic is the same that guides initiatives of digital sovereignty in several countries: who controls the data center controls the conditions of use of the models.

Agentic AI transforms demand for computing

The AI — systems that perform multiple-stage tasks with little human intervention — was one of the axes of the event. This type of application consumes continuous, non-time processing: instead of a heavy and sporadic training, spending focuses on inference running all the time.

This displaces the demand for infrastructure beyond training and explains why NVIDIA insists on the speech of "IA in scale". The theme already appeared in the company's strategy of covering the whole cycle, from training cloud to production, as in the agreement with CoreWeave. For European partners, the message is that the bottleneck is no longer the model and has become the capacity to serve agents in volume.

NVIDIA GTC Berlin: what changes to devs in Brazil?

Not very directly — and quite in the medium term. The announcement is European, but signals where NVIDIA is placing capacity and which partners gain supply priority. Brazilian developers access this infrastructure mostly via cloud and resellers, so changes in regional availability are delayed.

The most concrete effect is regulatory and expectation. While Brazil discusses its own AI milestone, the European debate on data residence tends to influence the demands of corporate and public sector customers here. Companies that sell to these buyers should start answering questions about where the data is processed.

What the NVIDIA announcement still does not solve

The material does not detail total values of European investment, delivery schedules or the full list of partners involved. There is also no indication of how much capacity will actually be installed on European soil and how much will be accessed by cloud contracts.

The continent's structural restrictions remain: energy availability, licensing of data centers and specialized labor. AI sovereignty depends on these factors as much as on GPUs — and no platform announcement solves this on its own.

Frequently Asked Questions about NVIDIA GTC Berlin

What is NVIDIA's GTC?

It is the technical conference of NVIDIA, where the company presents platforms, chips and partnerships. The Berlin edition takes the format to Europe, focusing on AI computing, AI agency and data sovereignty.

What does "sovereign AI" mean in practice?

It means keeping data and models processed under the laws and jurisdiction of a country or block, with infrastructure installed in the region. In Europe, this translates into a requirement for local data centers and data processing rules.

Does this affect developers in Brazil?

Indirectly. The ad sets supply and capacity expansion priorities, which can change cloud availability and computing price. It also anticipates data residency requirements that corporate customers tend to charge from Brazilian suppliers.