Computex 2026: Nvidia's 'Vera Rubin' Enters Production, Powering Next-Gen AI Data Centres

TAIPEI, June 1. At Computex 2026, one of the world's biggest technology shows, Nvidia made a major announcement for the world of artificial intelligence. Chief executive Jensen Huang revealed that the company's next-generation server platform, Vera Rubin, has now entered full production. The platform will form the backbone of the next generation of AI data centres, with shipments set to begin in the fall of 2026. Held in Taiwan's capital Taipei, the show is where the world's leading technology firms unveil their newest products, and this year Nvidia's announcements drew the most attention.

Vera Rubin: the power of new technology

The Vera Rubin platform has two key components—a new 'Vera' CPU, an 88-core design based on Nvidia's Olympus core, and a new GPU built on the 'Rubin' architecture. Both chips are manufactured on TSMC's 3nm process, a generation ahead of the N4 process used in the current Grace Blackwell duo. Nvidia claims the platform delivers a 10x gain in agent throughput over Grace Blackwell, dramatically improving the cost and speed of running AI models. The platform is named after the renowned American astronomer Vera Rubin, a pioneer in the study of dark matter—Nvidia frequently names its architectures after great scientists.

A flurry of awards

Nvidia's 'Vera Rubin NVL72' system was the centre of attention at Computex. It was named "Best Choice of the Year," and won a Golden Award along with the Sustainable Tech Special Award. The recognition underlines not just the chip's performance but its superiority on energy efficiency—an increasingly critical concern in the era of massive AI data centres. As AI models grow ever larger, the electricity needed to run them has become a serious environmental and economic issue, and Vera Rubin's efficiency on this front is seen as a major achievement.

RTX Spark for the PC

Beyond Vera Rubin, Nvidia also unveiled a new superchip for Windows PCs in the era of personal AI agents. Built in collaboration with MediaTek, the 'NVIDIA RTX Spark' is designed for AI, creation and gaming. The company also announced DLSS 4.5 and new GeForce RTX updates, making clear that Nvidia wants to keep its grip on the consumer market as well as the data centre. The move shows the company wants to bring AI not just to vast servers but right onto the desktops and laptops of everyday users.

Jensen Huang's hint

In his keynote, Jensen Huang offered an exciting glimpse of the rest of 2026. "The second half of this year is going to be very, very busy with Grace Blackwell, Vera Rubin, and we have a surprise new product that we haven't told anyone about yet," he said. He also laid out a three-generation roadmap of RTX Spark for laptops and desktop PCs, referencing the 'Rosa Feynman' architecture that follows Rubin. Huang, on stage in his trademark black leather jacket, is a figure whose presentations are always watched with great anticipation in the tech world.

Impact on the AI industry

Nvidia's announcements come at a time when companies worldwide are racing for the vast computing power needed to train and run AI models. Vera Rubin's 10x throughput and improved energy efficiency could lower the cost of large language models, benefiting everyone from cloud providers to startups. Nvidia is already the leader in the AI chip market, and this move further strengthens its lead. The company's market value has soared to historic highs in recent years on the back of the AI revolution, and Vera Rubin is seen as the key to sustaining that momentum.

What's next

With Vera Rubin shipping in the fall of 2026, the world's largest data centres will begin adopting the technology over the coming months. Huang's hint of a "surprise product" has already triggered a wave of industry speculation. Pressure on rivals such as AMD and Intel will grow, and the months ahead will determine how Nvidia maintains its dominance in the race for AI hardware. Meanwhile, analysts will also watch how manufacturers like TSMC handle the supply of these advanced chips amid surging demand.

Why it matters for India

For India, Nvidia's roadmap carries particular significance as the country accelerates its own AI ambitions. Indian technology firms, startups and government-backed initiatives are racing to build domestic data-centre capacity and sovereign AI capabilities, and chips like Vera Rubin sit at the heart of that effort. Major cloud providers operating in India depend on Nvidia hardware to power services used by millions, from enterprise software to consumer apps. Greater efficiency and throughput could lower the cost of training and deploying AI models tailored to Indian languages and use cases, potentially democratizing access for smaller players. As global demand for these chips far outstrips supply, securing allocation has become a strategic priority, and India's growing appetite for AI infrastructure means developments at events like Computex are watched closely by policymakers and industry leaders alike.