Nvidia is preparing to launch its new Rubin architecture-based GPUs six months earlier than previously estimated. According to new reports, the first GPUs developed with the Rubin architecture will be introduced in the third quarter of 2025 and will feature HBM4 memory and TSMC’s 3nm manufacturing process. Here are the details

Nvidia Rubin GPUs coming 6 months earlier than expected

Nvidia is currently offering Blackwell architecture-based GPUs to commercial customers for use in AI applications. The highly anticipated Rubin architecture is currently waiting in line to replace this series. Rubin GPUs, which were previously expected to be released in mid-2026, are planned to be introduced in the third quarter (Q3) of 2025. So why the early launch date?

According to the information we received, the decision to launch early may have been taken to respond to the growing demand in the AI market. The new Rubin GPUs will be launched primarily at the enterprise level. The release date of Nvidia’s consumer Rubin cards is not yet clear. Meanwhile, the Blackwell-based RTX 50 series, which will be released just ahead of Rubin in the consumer market, may be announced at CES 2025.

Rubin architecture, HBM4 (High Bandwidth Memory 4) technology and 3nm node structure will strengthen Nvidia’s hand in the market. Do you think this rapid transition will help Nvidia maintain its market superiority? What changes can Rubin architecture lead to in the world of gaming and AI? Don’t forget to share your comments with us.