Despite AMD’s introduction of a new generation of Navi family graphics cards with 7nm process and RDNA architecture, NVIDIA’s position in the high-end GPU market remains unshakeable, with the 12nm process’s Turing Turing graphics card still gaining the upper hand in performance and energy efficiency.
The strength of the 12nm GPU has kept NVIDIA conservative in the 7nm GPU, but in 2020 NVIDIA will launch the 7nm ampere family GPU chip, as soon as next year’s Q1 quarter, compared to the current service of “Turing”, One of the major architectural changes is a significant improvement in ray-tracing performance and a more powerful grate processing capability to better handle 3D scenes.
After the Ampere Amp GPU? Previously, it was rumoured that NVIDIA’s next-generation GPU, codenamed Hopper, came from the name of Ms. Grace Hopper, the mother of the compilation language, and was considered one of the earliest programmers in computer history to lead the development of the programming language Cobol.
According to information released by the U.S. Trademark Office, NVIDIA has officially obtained the Hopper trademark application on December 4, meaning that the rumored Hopper GPU has at least a bit of credibility.
Given the reality, it’s too early to say the details of the Hopper GPU, but previous tweeters have said that the HOPper generation’s GPU has changed dramatically and will introduce MCM’s multi-chip connected GPU design to create a monster-grade super core. This will be useful for turning on real-time ray tracing at 4K resolution, full height effects, and creating accelerators, professional cards, and more.
In addition, Hopper will use more advanced technology in the process process, and may upgrade the 6nm or even 5nm process.