NVIDIA Accelerates AI Chip Development: Rubin GPUs and Vera CPUs Set for Early Debut

NVIDIA's next-gen Rubin GPU, Vera CPU rumors: no delays, new chips are being moved up

NVIDIA's next-gen Rubin GPU and Vera CPUs will reportedly tape-out at TSMC in June, sampled in September, on TSMC N3P and CoWoS-L advanced packaging. As an Amazon Associate, we earn from qualifying purchases. TweakTown may also earn commissions from other affiliate partners at no extra cost to you. NVIDIA's next-generation Rubin GPU and Vera CPU chips will reportedly finalize tape-out at TSMC in June, and will enter trial production with sample chips provided in September at the earliest according to the latest reports. In a new post on X by financial analyst Dan Nystedt on X, we're hearing that TSMC will be making NVIDIA's next-gen Rubin AI GPUs on its N3P process node and will be using CoWoS-L advanced packaging. NVIDIA's new Rubin will be the company's first chiplet design for a GPU, after using traditional monolithic designs up until now (Hopper, Blackwell, etc are all monolithic large GPUs). Rubin will require even more advanced packaging capacity from TSMC with a 4x reticle size, which is a chunky piece of silicon compared to Blackwell and its 3.3x reticle size. The I/O die on NVIDIA's next-gen Rubin GPUs will also be in the package, fabbed on TSMC's N5B process node. NVIDIA's new Rubin AI GPUs are expected to enter mass production in early 2026, so we should expect much more news at CES 2026 and then a possible launch at GTC 2026 (NVIDIA's in-house GPU Technology Conference). Ctee reports that the supply chain has said that NVIDIA's new chip development schedule is "smoother than before", too. NVIDIA's next-gen Rubin AI GPUs will be using the latest bleeding-edge HBM4 memory, with SK hynix shipping out the world's first 12-layer HBM4 samples to its customers not too long ago. NVIDIA's massive new Vera Rubin NVL576 AI server that will be using 576 x Rubin R100 AI GPUs, and an insane 12672-core, 25344-thread CPU based on its new Vera architecture, all coming in 2026.

NVIDIA's Next-Gen Rubin AI Accelerators To Enter The Market as Soon As September, Just Six Months After Blackwell Ultra

NVIDIA has apparently boosted its product roadmap, as it is now revealed that Rubin AI chips and Vera CPUs are expected to debut as early as September. There's no stopping Team Green when it comes to the pace of innovation the firm has managed to achieve in the past few years, part of which has to be attributed to how quickly AI is evolving. NVIDIA has officially declared it to be on a "yearly" production cadence, which means that every architecture will be released to the market after 12 months; however, this schedule appears to be "halved", as the difference in schedules between Blackwell Ultra and Rubin is now estimated to be just six months. A report by Ctee now states that Rubin GPUs and Vera CPUs are expected to hit sampling as soon as September, with tape-out already initiated. For a quick rundown on what to expect with NVIDIA's Rubin, the firm will utilize the next-generation HBM4 chips to power its R100 GPUs, which are said to be a significant upgrade from the modern-day HBM3E standard. Team Green will also adopt TSMC's 3nm (N3P) process and CoWoS-L packaging, which means that Rubin will adopt newer industry standards that will likely take performance to greater levels. More importantly, Rubin will adopt a chiplet design, which will be a first-of-a-kind implementation from NVIDIA, and will use a 4x reticle design (versus 3.3x of Blackwell). NVIDIA's core focus with Rubin GPUs is pursuing incredible performance/watt figures, given that the growing power needs of data centers have compelled Team Green to adopt a sustainable development path. The firm will also replace its renowned Grace CPUs with the advanced Vera lineup, which is said to be based on ARM's next-generation cores and will likely come with a significant improvement in generational performance. NVIDIA's hardware stack will evolve into the next generation with the Rubin/Vera combo, and we'll likely see a massive boost in computational power. While the Rubin release certainly shows optimism, there's always a question about how NVIDIA would release independent architectures in such short intervals, given that the supply chain gets so little time to adopt newer frameworks. We have always seen issues emerge with the initial batch of AI products from NVIDIA, whether Blackwell or Blackwell Ultra, so it will be interesting to see how Team Green sorts out Vera/Rubin, lineups that have their hardware reworked from the ground up.