NVIDIA CEO Huang Renxun: Putting effort into AI does not mean neglecting game players, RTX 4060 Ti is suitable for 8GB graphics memory
04 Jun 2023
0 Comments
On June 5th, it was reported that during the recently held Taipei International Computer Show (Computex 2023), NVIDIA CEO Huang Renxun gave an interview to foreign media PC World.
The reporter asked, "Will current AI technology, like a newborn baby, make game players feel like they have been left out.
Huang Renxun joked that he "seemed to have been offended" and subsequently responded as follows:
“Firstly, NVIDIA's RTX graphics card is specifically designed for gaming players, with the most important technology being AI. Without AI technology, real-time ray tracing cannot be performed, and Nvidia's first AI project was also used for deep learning in DLSS. Deep learning is currently the backbone of RTX graphics cards.
“Firstly, NVIDIA's RTX graphics card is specifically designed for gaming players, with the most important technology being AI. Without AI technology, real-time ray tracing cannot be performed, and Nvidia's first AI project was also used for deep learning in DLSS. Deep learning is currently the backbone of RTX graphics cards.
Secondly, NVIDIA's first landing project of generative AI technology, Avatar Cloud Engine (ACE), can leverage AI driven natural language interaction to bring a more intelligent interactive experience to NPCs in games, thereby changing the game.”
It was learned from the interview that Huang Renxun also briefly mentioned the recently launched RTX 4060 Ti, which starts at 3199 yuan but only comes with 8GB of graphics memory. Therefore, many game players expressed dissatisfaction.
Huang Renxun responded that frame caching (graphics memory) is not computer memory, and how to manage caching is a big issue. Although larger graphics memory is better, it still requires a lot of balance. Huang Renxun compared it to an equation or "kung fu" - a balance needs to be made between cost, power consumption, and core, and stated that 8GB of graphics memory is the most accurate capacity.
