On June 2, NVIDIA CEO Jensen Huang delivered a keynote speech at the National Taiwan University (NTU) Sports Center, where he unveiled the new Rubin architecture. He highlighted NVIDIA’s accelerated rollout of new architectures.
Huang discussed NVIDIA’s upcoming Blackwell Ultra GPU and hinted at its continued upgrades. He then introduced the Rubin architecture, which will succeed Blackwell. The Rubin GPU will feature 8 HBM4 chips, while the Rubin Ultra GPU will include 12 HBM4 chips. The new architecture is named after American astronomer Vera Rubin, known for her groundbreaking work on dark matter and galaxy rotation rates.
Despite the recent launch of the Blackwell platform, NVIDIA is rapidly advancing its roadmap. According to Huang, the Rubin GPU will join the R series products and is slated for mass production in the fourth quarter of 2025. The Rubin GPU and its platform are expected to debut in 2026, with the Ultra version following in 2027. Both versions will utilize HBM4 memory.
Wccftech reports that the Rubin GPU will adopt a 4x reticle design, employ TSMC’s CoWoS-L packaging technology, and use the N3 process. Additionally, NVIDIA plans to upgrade from the current HBM3e in its B100 GPU to HBM4 as soon as HBM4 solutions become widely available by late 2025.
Have you read?
Richest Billionaire Investors.
Billionaire Winners.
Billionaire Losers.
Best Business Schools.
Best Hotel Schools.
Add CEOWORLD magazine to your Google News feed.
Follow CEOWORLD magazine headlines on: Google News, LinkedIn, Twitter, and Facebook.
Copyright 2024 The CEOWORLD magazine. All rights reserved. This material (and any extract from it) must not be copied, redistributed or placed on any website, without CEOWORLD magazine' prior written consent. For media queries, please contact: info@ceoworld.biz
