‘Next generation of Nvidia RTX video cards will have a monolithic chip design’

YouTuber Moore’s Law Is Dead claims that the next generation of Nvidia RTX GPUs will have a monolithic chip design. The claims are consistent with those of YouTuber RedGamingTech who states that the RTX 50 GPUs will be twice as fast as the RTX 40 GPUs.

Moore’s Law Is Dead claims that Nvidia does not want to take any risks with the design of video cards and has therefore once again opted for a monolithic chip design. According to the YouTuber, this production method costs more compared to that of chiplets, but according to Nvidia it is more reliable. Nvidia would also like to focus more on the development of AI software in the coming years to keep its video cards performing.

Moore’s Law Is Dead claims are in line with those of YouTuber RedGamingTech. It also claims that the consumer versions of the next generation of GPUs will have a monolithic chip design and that the underlying architecture will be extensively tinkered with. Nvidia would focus on ray tracing and path tracing, but would also work on a denoise accelerator.

RedGamingTech believes that the next generation of Nvidia video cards use PCIe 5.0 and that the GPUs will be produced on a 3nm chip process from TSMC. GDDR7 memory will also be used. This was also previously claimed by the editors of the Chinese tech website MyDrivers. According to RedGamingTech, RTX 50 GPUs are on average twice as fast as RTX 40 GPUs. Although, according to the YouTuber, these last claims should not be taken too literally.