After the release of the new generation of GeForce RTX 4090 GPU, Nvidia CEO Jen-Hsun Huang confirmed to reporters that the company is canceling its NVLink bridge design, and canceling the NVLink connector can free up more space, thereby Handle more AI operations.
"The reason we got rid of [NVLink] was that we needed the I/O to do other things, so we used this area to cram as much AI processor as possible," Jen-Hsun Huang explained the reason for getting rid of NVLink.
For gamers, creators, and workstation users who want to use the latest RTX 4000 series for multi-card firefights, Nvidia can also achieve similar support with PCIe Gen 5 after NVLink is removed.
"Because Ada [Lovelace] is based on PCIe Gen 5, we now have the ability to achieve point-to-point transfers that are fast enough, which is a good deal," added Jen-Hsun Huang.
Compared to PCIe 4, PCIe 5 doubles the bandwidth with direct access to the CPU lanes, providing extraordinary speed benefits for storage, GPU, and memory.
In addition, the RTX 4090 and RTX 4080 graphics cards only have HDMI 2.1 and DP 1.4a interfaces and do not have the latest DP 2.0 interface. For this, NVIDIA officially explained in the community's Q&A, saying that the DP 1.4 interface can now support 8K 60Hz output.
In view of the high price of the RTX 4090 when it was launched, Huang Renxun also announced that "Moore's Law is dead, and the foundry price of 12-inch wafers is much more expensive than before, and the idea of chip prices falling is a thing of the past."
He said prices would rise across the supply chain, and the huge generational increase in performance made it impossible to meet the "same performance for half the price" condition, or twice the performance for the same money, like the previous generation RTX 3000 series.
"Nvidia's $899 GPU or $1599 GPU, we've got much better performance with Ada Lovelace at the same price point compared to a year ago, two years ago."
The NVIDIA RTX 4090 flagship graphics card, equipped with 16384 CUDA + 24GB video memory, has a suggested price of 12,999 yuan and will be available on October 12.
According to reports, the RTX 4090 is the fastest gaming GPU in the world. In terms of performance, Nvidia claims that the RTX 4090 with DLSS 3 is up to 4x faster than the previous-generation RTX 3090 Ti with DLSS 2 in full ray-traced games. In modern games, the RTX 4090 delivers up to 2x faster performance while maintaining the same 450W power consumption.
With 76 billion transistors, 16,384 CUDA cores, and 24 GB of high-speed Micron GDDR6X memory, the RTX 4090 consistently runs at over 100 FPS in games at 4K resolution.
Top graphics card suppliers such as ASUS, Colorful, Gainward, GALAXY, GIGABYTE, Inzone, MSI, and Zotac will launch GeForce RTX 4090 standard and overclocked versions in China. Acer, Alienware, Asus, Dell, HP, Lenovo and MSI will also launch OEM products. In addition, NVIDIA will also launch the FE version of the RTX 4090 in limited quantities.
0 Comments