NVIDIA GTC Berlin Conference Opens Today: Jensen Huang to Deliver Keynote Tomorrow; the 'AI Factory' Faces Its First Stress Test

Once enterprises deploy Agents at scale, inference workloads will explode exponentially. ## Why It Matters: A Public 'Stress Test' Viewing GTC Berlin through the lens of the current 'accountability phase' reveals its significance far beyond that of a typical technology launch. First, product generations dictate supply cadence, while customer deployments determine revenue realization. NVIDIA’s Q2 FY2027 (ended July 26, 2026) reported total revenue of $96.2 billion, with data center revenue reaching $89 billion—a 117% year-on-year increase—accounting for over 92% of total revenue. For the prior full fiscal year (FY2026), data center revenue totaled $193.7 billion, up 68% year-on-year. While the core business remains robust, the 'second derivative' of growth is now under renewed scrutiny. Second, the 'AI factory' concept itself is a direct response to this accountability imperative. When customers no longer buy individual GPUs but entire computing infrastructure stacks, evaluation criteria shift—from 'how much performance improvement?' to 'cost per token, system utilization, energy efficiency, and whether applications can sustainably generate revenue.' Third, the race between inference efficiency and scale remains unresolved. Currently, two countervailing forces shape inference: on one hand, inference efficiency continues improving, potentially reducing compute resources required per task; on the other, lower costs may drive broader adoption and larger-scale deployment. What ultimately determines whether inference compute can sustain monetization is not lab-based technical demos—but real-world ROI for enterprise customers. Only when Agents enter genuine business workflows—customer service, software development, enterprise operations, and scientific research—can foundational inference demand meaningfully expand across a wider customer base. Jensen Huang himself, speaking last week in Berlin amid market volatility, remarked: 'This too shall pass—and for a very simple reason: AI will bring transformation. It will positively impact every industry and every country in profound ways. There is no doubt this impact will endure—and we are only at the beginning of this journey.' These words offer reassurance—but markets demand execution. ## Impact: Three Transmission Channels For the supply chain: Humanoid robots have entered a critical mass-production validation phase. Tesla’s Optimus design is nearing finalization, and mass-production expectations are rising; domestic automotive component production lines have already begun incorporating humanoid robots into routine operations. GTC’s Physical AI and robotics agenda will directly influence order forecasts for components including ball screws, gear reducers, servo motors, joint modules, and dexterous hands. For Europe: Hosting GTC in Berlin itself sends a strong signal. Europe is becoming a pivotal market for NVIDIA’s next-phase AI ecosystem expansion. Under the 'sovereign AI' agenda, European governments are willing to fund domestic AI infrastructure. The allocation of NVIDIA’s newly announced $2 billion investment will determine who secures the next piece of Europe’s AI landscape. For everyday people: Platforms like Vera Rubin aim ultimately to reduce inference costs. As inference becomes cheaper, AI applications can transition from 'novelty' to 'everyday utility,' and Agents can truly enter customer service, office productivity, and research environments.

Deutsche Telekom and NVIDIA Collaborate to Build an AI Superfactory

News Summary Deutsche Telekom and NVIDIA plan to jointly build an industrial AI computing center in Germany. The data center will deploy 10,000 NVIDIA AI chips. This project constitutes the first phase of the EU’s planned 100,000-chip AI superfactory. Deutsche Telekom is leveraging this initiative for strategic transformation, doubling down on the AI infrastructure sector. Full Text Deutsche Telekom and U.S. chipmaker NVIDIA announced plans to jointly construct a high-performance computing center in Germany focused on industrial artificial intelligence (AI) applications. The two companies officially unveiled the collaboration last Friday. 01 The Initiative Aligns with Germany’s National Strategy Deutsche Telekom Group CEO Timotheus Höttges emphasized: 'We must seize the opportunities presented by AI to fundamentally transform German industry and secure our leadership position in global technological competition.' NVIDIA CEO Jensen Huang arrived in Berlin and met with German Chancellor Friedrich Merz on Friday. Following the meeting, Chancellor Merz stated: 'Investments in strategic AI infrastructure are vital to enhancing our nation’s innovative capacity.' He also commended the computing center project and NVIDIA’s commitment, calling the partnership 'a potentially pivotal step toward Germany’s digital sovereignty and economic future.' While referencing Germany’s partners, he did not explicitly name Deutsche Telekom. 02 Project Implementation Details According to Deutsche Telekom, the company will be responsible for constructing, operating, and commercializing the data center—but has not disclosed the specific investment amount. The industrial AI cloud platform is scheduled to be completed no later than 2026. The data center, operated by Deutsche Telekom, will be equipped with 10,000 NVIDIA AI chips based on the Blackwell architecture, with software built upon NVIDIA’s CUDA platform.

Germany's Merz urges AI investment ahead of Nvidia deal

3 days ago

A German business coalition proposes €4.1 trillion investment in AI, energy, and defense by 2035, urging reforms to support sectors like chemicals. Chancellor Merz will present the plan ahead of a Berlin investment summit.

reuters

2026-10-19

Germany's Merz urges AI investment ahead of Nvidia deal

Germany's Merz urges AI investment ahead of Nvidia deal

FAQ