A revised set of specifications for the Nvidia GeForce RTX 4090 has been published by a well-known Twitter leaker. According to Kopite7kimi, Nvidia’s next-generation Ada Lovelace architectural flagship will include 126 Streaming Multiprocessors (SMs).
The value is significantly lower than the previously reported and widely anticipated figure of around 140 SMs. Please take this information with a grain of salt, as with all leaks.
Kopite7kimi felt secure enough after giving the preceding piece of information to present a comprehensive set of specs expected for the RTX 4090.
Despite the Ada Lovelace architecture Nvidia GeForce RTX 4090’s lower wattage and fewer SMs / CUDA cores, Kopite7kimi expects it to produce twice the 3D acceleration performance of its predecessor
With the reduction in the claimed number of SMs for the RTX 4090, it’s encouraging to see the power needs to be reduced to 450W. Previous speculations claimed that the power usage would be as high as 600W.
We reported over the weekend on Kopite7kimi’s information that the GeForce RTX 40 series will arrive earlier in 2022 than expected. Nvidia’s next-generation consumer graphics cards will be released in early Q3, according to a tech Twitter leaker. The launch window has now been narrowed down to “mid-July,” according to Kopite7kimi.
Due to the lack of a major PC gaming or technology event in mid-July, it appears like Nvidia will have to organize its special launch event. However, enthusiasts are less concerned with the scale, location, or other elements of the event, preferring instead that the products function well, arrive swiftly, and are reasonably priced.
According to Kopite7kimi, we are presently travelling through mid-May, which means the introduction of the GeForce RTX 40 series is only two months away. There are still many unanswered issues around the nature of the launch, the initial launch lineup, full specifications, availability, and cost.
Some of these questions may be definitively answered during Computex next week if we’re lucky. For the time being, all we can do is wait and watch.
also read:
