Nvidia has increased its efforts to develop its RTX 50 series, which is scheduled to launch in the coming year. Certain technical aspects that are part of the Nvidia RTX 50 family, which is believed to be launched using Blackwell, have been revealed. Blackwell design, have come to light.
The features from the Nvidia RTX 50 family become apparent
According to leaked information, Nvidia will include transistors manufactured by TSMC with 3nm technology on its next generation graphics cards. The acceleration of 3nm production will begin in the 4th quarter in 2022. TSMC has announced that its production costs rose by 25 percent in comparison the 5nm process. It’s important to note that the Nvidia CEO, who is based in Taiwan, visited Taiwan to discuss with TSMC about securing the early 3nm technology for future-generation graphics cards.
The company will be launching with a monolith design within the RTX 50 series. A renowned analyst Kopite7kimi said that the Blackwell architecture was designed for the user as well as the server.
Based on the information released in the report by RedGamingTech, Nvidia will come with major changes to the CUDA architecture of Blackwell graphic cards. According to the report, Nvidia will enhance the hardware that is used for visual enhancement as well as RayTracing.
Samsung is believed for its work on new GDDR6W DRAM, which will give even more performance and larger capacities. However, the use of GDDR7 memory technology is yet another important information leak on this RTX 50 series. It is important to know that data transfer speeds of up to 36 Gbps are possible with GDDR7 technology.
It is believed about the Blackwell GeForce the RTX series utilizes PCIe Gen 5 interface and provides a faster clock speed than 3GHz. Additionally, the new generation of graphics cards are believed to be two times more efficient than Ada GPUs when it comes to performance.