Friday, December 9, 2022
HomeDigit NewsThe core of the Radeon RX 7900 is designed to operate at...

The core of the Radeon RX 7900 is designed to operate at frequencies up to 3 GHz. But it is not yet clear whether such a frequency will be achievable in reality.

The core of the Radeon RX 7900 is designed to operate at frequencies up to 3 GHz. But it is not yet clear whether such a frequency will be achievable in reality.

AMD confused users a little with their statements

Video cards Radeon RX 7900, contrary to rumors, cannot boast of any ultra-high GPU frequencies, if we talk about the declared characteristics. What frequencies will actually be, we will find out from the first reviews, but now it has become known that the new Navi 31 GPU is designed to operate at frequencies up to 3 GHz.

This was allegedly told by AMD itself, although it did not say a word about this at the official presentation. In addition, there is information that with a preliminary driver that AMD sent out to manufacturers, the overclocking potential of the GPU is limited to only 3%, but with the final figure they can increase. In particular, Asus for their video card TUF frequency does not yet indicate, which may mean that she simply has not approved them yet.

Whether it is possible to expect that in reality in games the GPU of the new AMD adapters will operate at a frequency of about 3 GHz is not yet clear, but for some reason the company mentioned this.

By the way, AMD also noted that for its new adapters it is incorrect to indicate the number of stream processors that all the media indicated in their materials. Recall that for the RX 7900 XTX, this is 12,288 stream processors. If you go to the company’s website, you can see an indication of the presence of 6144 stream processors, that is, half as many. The fact is that the RDNA 3 architecture relies on the Dual Issue design, which means the execution of two FP32 arithmetic commands at the same time, which is why we can talk about supposedly 12,288 FP32 blocks or stream processors. However, AMD still wants the data to be correct. By the way, Nvidia also implemented something similar when switching to the Ampere architecture, due to which the number of CUDA cores relative to RTX 20 increased much more than performance increased.

- Advertisement -
MOBILE

- Advertisment -

- Advertisment -spot_imgspot_imgspot_img

Auto

Two Wheeler

BUSINESS AND FINANCE

Digit News

spot_img