What process is used for rtx50 graphics card?

NVIDIA’s next graphics card plan must be RTX50, so many people don’t know about it. Although there is currently a lot of exposed information on the market, there are several problems, or it is complicated and difficult to understand. Or it just looks complicated and people are too lazy to read it. The editor has summarized a few questions that everyone is looking forward to. I believe the most critical question is what process is used for the RTX50 graphics card?

According to foreign media reports, the GB200 series GPU used in the Nvidia RTX 50 series graphics card will use TSMC’s 3nm process.

Judging from the latest details exposed, the CUDA cores of the flagship product RTX 5090, codenamed GB202, have increased by 50%, with the total number reaching 24,576. That’s more than 8 times more cores than the Turing-based RTX 2080.

In addition, the memory bandwidth of the new flagship card will also be increased by 52% by using 32Gbps GDDR7, which is higher than the GDDR6X of the current RTX 4090.

The above is the detailed content of What process is used for rtx50 graphics card?. For more information, please follow other related articles on the PHP Chinese website!