NVIDIA has revealed more details of its next-generation AI Data Center solution, Rosa Feynman, built using 3D Die-Stacking.
NVIDIA Goes With 3D Die Stacking & Custom HBM Solution For Its Next-gen Rosa Feynman AI Chips
NVIDIA's Feynman GPU architecture was confirmed back at GTC 2025. During the announcement, NVIDIA listed the Feynman GPU with next-gen HBM, Vera CPU, and several other connectivity chips that make up the foundation of the AI data center.
Well, this year at GTC 2026, NVIDIA is revealing more information and some big changes. First up, NVIDIA is now confirming that its Feynman GPUs will adopt 3D Die Stacking technology. With 3D Die Stacking, we could very well see the first use of 3D stacked GPU dies from NVIDIA. It also looks like NVIDIA will utilize Intel as a Foundry partner and leverage their advanced packaging technologies, such as EMIB, to produce Feynman chips.
The other exciting announcement is that instead of next-gen HBM, NVIDIA is now listing custom HBM tech for its Feynman GPUs. With Rubin leveraging HBM4 and Rubin Ultra leveraging HBM4E, it looks like NVIDIA's solution might be either a custom or heightened version of HBM4E or a custom HBM5 solution, which makes them stand out from the standard HBM5 offerings.
NVIDIA is also confirming the name of its next-gen Data Center CPU architecture. Feynman isn't going to rely on Vera; it is going to use a brand new CPU called Rosa, which is named after American physicist & a Nobel Prize winner, Rosalyn Sussman. No details are mentioned yet, but given NVIDIA's trajectory, we can expect some big improvements.
Besides this, NVIDIA will continue to roll out a full spectrum of chips for its AI platforms, such as BlueField-5, NVLink 8 CPO, Spectrum 7 204T, CPO, and CX10. As expected, NVIDIA's Rosa Feynman solutions will be arriving in 2028.
NVIDIA Data Center / AI GPU Roadmap
| GPU Codename | Feynman | Rubin (Ultra) | Rubin | Blackwell (Ultra) | Blackwell | Hopper | Ampere | Volta | Pascal |
|---|---|---|---|---|---|---|---|---|---|
| GPU Family | GF200? | GR300? | GR200? | GB300 | GB200/GB100 | GH200/GH100 | GA100 | GV100 | GP100 |
| GPU SKU | F200? | R300? | R200? | B300 | B100/B200 | H100/H200 | A100 | V100 | P100 |
| Process Tech | TSMC A16? | TSMC N2P? | TSMC N3P? | TSMC 4NP | TSMC 4NP | TSMC 5nm | TSMC 7nm | TSMC 12nm | TSMC 16nm |
| CPU | Rosa | Vera | Vera | Grace | Grace | Grace | N/A | N/A | N/A |
| Memory | HBM4e/HBM5? | HBM4 | HBM4 | HBM3e | HBM3e | HBM2e/HBM3/HBM3e | HBM2e | HBM2 | HBM2 |
| Launch | 2028 | 2027 | 2026 | 2025 | 2024 | 2022-2024 | 2020-2022 | 2018 | 2016 |
About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.
Follow Wccftech on Google to get more of our news coverage in your feeds.
