Industry News

The NVIDIA RTX 50 series graphics card is still manufactured by TSMC: 2 advantages are not solved

Views : 24
Update time : 2023-06-02 12:01:08
        On June 1, Nvidia's RTX 30 series GPU chips were OEM by Samsung's 8nm process. The current RTX 40 series turned to TSMC's 4N process (5nm customized version). CEO Huang Renxun complained many times last year. TSMC foundry prices are too expensive, but they can't give up, and the next generation of chips will continue to be used.
 
 
        CEO Huang Renxun said in an interview this afternoon that he will meet with executives from TSMC, Foxconn and other companies this Friday. He also mentioned that Nvidia's next generation chips will continue to use TSMC foundry. Not only will it purchase chips produced by local factories, but some orders will also be handed over to TSMC's Arizona plant for production.
Although the foundry price of TSMC's advanced technology is the highest in the industry, Huang Renxun also mentioned two advantages of TSMC, one is huge production capacity and the other is incredible agility.
        In other words, although Nvidia will also dislike TSMC's high-cost foundry, TSMC's process yield and production capacity are guaranteed, and it is extremely flexible, and these advantages cannot be replaced by other foundries. You know, even if Samsung OEM RTX 30 gaming card GPU with 8nm process, NVIDIA's A100 acceleration chip was also OEM.
        As for the next-generation product mentioned by Huang Renxun, although he did not specify what it was, the next-generation GPU had already been dug up before. This GPU was codenamed Blackwell, who was the first black academician of the National Academy of Sciences.
        Among them, Blackwell's RTX 50 series for gaming cards will be designed with a single GPU chip, and Blackwell for HPC and AI will be designed with MCM multi-chip. NVIDIA is also familiar with this. Grace Hooper is now packaged with two cores CPU and GPU.
        What is less certain is the Blackwell process. A lot of news has pointed to TSMC 3nm before, but considering NVIDIA's habits, the 5nm process has only been used for one generation, and it is unlikely to switch to the higher-cost 3nm process. Blackwell chip in 2024 is likely to continue to use the 5nm process for gaming cards, and the 3nm process for Blackwell's AI and computing card versions.

 
Related News
Read More >>
How many chips does a car need? How many chips does a car need?
Sep .19.2024
Automotive chips can be divided into four types according to their functions: control (MCU and AI chips), power, sensors, and others (such as memory). The market is monopolized by international giants. The automotive chips people often talk about refer to
Position and Function of Main Automotive Sensors Position and Function of Main Automotive Sensors
Sep .18.2024
The function of the air flow sensor is to convert the amount of air inhaled into the engine into an electrical signal and provide it to the electronic control unit (ECU). It is the main basis for determining the basic fuel injection volume. Vane type: The
Chip: The increasingly intelligent electronic brain Chip: The increasingly intelligent electronic brain
Sep .14.2024
In this era of rapid technological development, we often marvel at how mobile phones can run various application software smoothly, how online classes can be free of lag and achieve zero latency, and how the functions of electronic devices are becoming mo
LDA100 Optocoupler: Outstanding Performance, Wide Applications LDA100 Optocoupler: Outstanding Performance, Wide Applications
Sep .13.2024
In terms of characteristics, LDA100 is outstanding. It offers AC and DC input versions for optional selection, enabling it to work stably in different power supply environments. The small 6-pin DIP package not only saves space but also facilitates install