According to reports, Nvidia is gearing up to launch its next generation of gaming graphics cards in 2022. The company's next-generation GeForce RTX 40 series GPUs are in the works and will be based on the Ada Lovelace architecture.
According to Wccftech, various reliable leakers have confirmed that Nvidia is utilising TSMC's 5mn process node for the new 40 series graphics cards coming in 2022. This has now been confirmed by information coming straight out of the Taiwanese factories where the graphics cards are being made.
- GPU shortage to continue throughout 2022, according to Nvidia
- Best graphics card: Get a GPU to power you to gaming bliss
Twitter user RetiredEngineer has shared some of the information originally posted on Digitimes:
"Nvidia's biennial GPU refresh coming in 2022, riding on metaverse and gaming. Following H100, based on Hopper architecture, using TSMC's 5nm + CoWoS, aimed at datacenter/AI, gaming GPU RTX40 series, based on Ada Lovelace architecture, will also tap TSMC's 5nm...."
— RetiredEngineer® (@chiakokhua) November 30, 2021
Being based on the Ada Lovelace architecture, these new GPUs are suggested to offer a number of improvements and innovations that should make them interesting.
The leaks suggest that the Nvidia GeForce RTX 4090 GPU - the flagship of the range - will be powered by Ada Lovelace AD102. That GPU will have 18432 CUDA Cores (a significant increase from the 10496 on the current RTX 3090). It could also have a clock speed as high as 2.5 GHz and as much as 92 TFLOPs of compute performance. All large jumps from the top end 30 series cards.
If these specs are accurate, there could be as much as a 150% performance jump - a similar leap that we saw between the RTX 2080 TI and the RTX 3090. In the real world that equated to around a 50% gaming performance improvement so it would certainly be interesting to see that again.
Nvidia's RTX 40 series graphics cards could go head-to-head with AMD's RDNA 3 based Radeon RX 7000 series graphics cards. But only time will tell how much of this is accurate.
from Pocket-lint https://ift.tt/3o37jt4
Comments
Post a Comment