A quarter that rewrote the record books

At the end of May, Nvidia announced its first quarter fiscal 2027 results — and the numbers are breathtaking. Revenue reached $81.6 billion, representing year-over-year growth of 85%. Net profit grew even more sharply, more than tripling to $58.3 billion.

The growth is primarily driven by data centers. The Data Center division posted record revenue of $75.2 billion — 92% more than a year ago. "The construction of AI factories — the largest infrastructure expansion in human history — is accelerating at an extraordinary pace," commented CEO Jensen Huang on the results. "Agentic AI is here, doing productive work, generating real value, and scaling rapidly across companies and industries."

And the optimism continues in the outlook: for the second quarter, Nvidia expects revenue around $91 billion. To put that in perspective — that's more than the annual GDP of Slovakia.

Vera Rubin: A processor built for agents

The biggest technological innovation of recent weeks is the Nvidia Vera Rubin platform. It includes both a new generation of GPU accelerators for training AI models and an entirely new Vera CPU processor — Nvidia's first processor designed specifically for agentic AI.

Why does this matter? Agentic AI — systems that plan, decide, and execute complex tasks on their own without constant human guidance — requires a different type of computing power than classic chatbots. According to Nvidia, Vera CPU offers "the highest number of instructions per clock in the world" — specifically 10 instructions per cycle. This means faster data processing that agents need for real-time decision-making.

"CPUs for agents are our new main growth engine," Huang declared at the Computex conference. And the numbers back him up — data centers now account for over 92% of the company's revenue.

RTX Spark: Nvidia takes on Intel and AMD

This year's Computex brought an even bigger surprise. Nvidia introduced RTX Spark — its own ARM processors for laptops and desktop computers, combining CPU and GPU on a single chip. This marks Nvidia's biggest entry into the personal computer market since the days when it only manufactured graphics cards.

The first wave of RTX Spark devices will arrive in the fall of 2026. Confirmed partners include Asus, Dell, HP, Lenovo, MSI, and Microsoft, which has already shown the Surface Laptop Ultra powered by Nvidia chips. A total of over 30 laptops and 10 desktops are in development.

What does this mean for the average user? RTX Spark promises up to 128 GB of unified memory, significantly longer battery life, and the ability to run large AI models locally — without the need for a cloud connection. For developers, designers, or anyone working with AI tools, this could mean a fundamental shift in productivity.

35 supercomputers for Europe — and the Czech Republic is part of it

At the ISC High Performance 2026 conference in Hamburg, Nvidia announced that a record 35 new AI supercomputers built on its technologies are being created across Europe. Together they will offer over 800 AI exaflops of performance and make cutting-edge infrastructure accessible to more than 3 million researchers in 23 countries.

The most significant projects include:

• MareNostrum 5 in Barcelona — an upgrade to NVIDIA GB300 chips will deliver up to 20 exaflops for AI training
• IT4LIA at Italy's CINECA — over 8,000 GPUs with 82 exaflops of performance
• HammerHAI in Stuttgart — Germany's first AI factory
• Blue Swan in Bavaria — 1,000 GPUs for science and industry

For Czech readers, it is crucial that the Czech AI Factory in Ostrava — which we have covered in detail at jarvis-ai.cz — is part of this European network. Czech researchers, startups, and companies thus gain access to the same infrastructure on which the largest European research projects run. And Nvidia is the common denominator of the vast majority of them — over 90% of European AI factories run on its technologies.

Halos: A safety shield for physical AI

While we hear about language models and chatbots every day, Nvidia is simultaneously building infrastructure for physical AI — robots, autonomous vehicles, and intelligent machines operating in the real world. And with that comes a fundamental challenge: safety.

In June, the company introduced Nvidia Halos for robotics — the world's first full-fledged safety system for physical AI. It is a software-hardware stack that ensures robots and autonomous systems function safely even in unpredictable situations. The first partner to deploy Halos is Agility Robotics, the maker of the humanoid robot Digit.

Why does this matter? Without safety standards, autonomous machines cannot legally operate alongside humans — whether in factories, warehouses, or on the streets. Halos is a step toward certification that will open the door to mass deployment of robots in industry.

What this means for Czech companies and developers

Nvidia is no longer just a hardware manufacturer. It is becoming a vertically integrated platform covering everything from chips to software to end-user devices. For Czech companies, this brings several practical implications:

Availability of computing power. With 35 new supercomputers in Europe (including the Czech node in Ostrava), even smaller companies and startups will have access to infrastructure that just a few years ago was reserved only for tech giants.

A new generation of computers. RTX Spark will bring laptops to the Czech Republic this fall that can run large language models locally. For developers, this means faster work with AI agents without cloud latency — and without monthly API fees.

Certified robotics. If Czech companies are considering deploying autonomous robots in manufacturing or logistics, the Halos standard gives them confidence that the technology will meet strict European safety regulations.

But it's not without risks

Amid all this "green wave," it's only fair to mention the flip side. Nvidia faces geopolitical restrictions — it cannot export its most powerful chips to China, which costs it billions in the outlook. Another uncertainty is the energy intensity of AI infrastructure. According to the UN, AI will consume up to 3% of the world's electricity by 2030 — and Nvidia, as the main hardware supplier, is at the center of this debate.

Yet the direction is clear. Nvidia is expanding into all key AI segments simultaneously — from cloud data centers to personal computers to safety systems for robots. And at least for 2026, it looks like every one of these directions is showing green.