Nvidia RTX 40-series will let you ramp up the power draw

A new collection of Nvidia GeForce RTX 40 series specs reveals an interesting feature: a configurable TGP. According to the leak, Nvidia’s best next-generation graphics cards will allow the user to trade higher power consumption for better performance, and all of this will be customized based on your needs rather than locking the GPU to a particular one. TGP.

There’s also another surprise: a popular GPU seems to be missing from the lineup.

The following is a summary for easy reading. NO NEWS.
-RTX 4090, PG136/139-SKU330, 16384FP32, 21Gbps 24G, 450W+.
-RTX 4080 16G, PG136/139-SKU360, 9728FP32, 23Gbps 16G, 340W+.
-RTX 4080 12G, PG141-SKU340/341, 7680FP32, 21Gbps 12G, 285W+.
I’m basically done with my work.

— kopite7kimi (@kopite7kimi) September 14, 2022

Kopite7kimi, a well-known leaker in the PC hardware space, shared a rundown of Nvidia GPU specs. There are no major updates there, although some changes can be noticed. More importantly, Kopite closed the collection with a short statement saying, “I’m done with my work.” This strongly implies that these specs could be the real deal Nvidia will be announcing soon.

We have equipped the flagship RTX 4090, based on AD102-300 GPU, with 16,384 CUDA cores and a boost clock up to 2,520MHz. The card comes with 24 GB of GDDR6X memory clocked at 21 Gbps, which equates to a huge bandwidth of 1 TB/s. The base TGP would be 450 watts.

It seems that Nvidia has changed its mind about the lineup, because instead of offering an RTX 4080 and RTX 4070, we may be getting two different versions of the RTX 4080.

The first model, with the AD103-300 GPU, has 16 GB of GDDR6X memory with a clock speed of 23 Gbps. It will also have 9,728 CUDA cores and a boost clock very close to the flagship (2.505 MHz).

This is followed by the AD104-400 based RTX 4080 12GB model with 7,680 CUDA cores and a higher boost clock up to 2,610MHz. The first version of the card will ship with a TGP of 340 watts, while the weaker model should have a more conservative 285 watts.

Interestingly, no RTX 4070 is listed, at least for now. Previous leaks indicated that the AD104-400 GPU would end up in an RTX 4070 Ti or an RTX 4070, but according to Kopite7kimi, Nvidia has decided to scrap that and release two different versions of the RTX 4080 instead. VideoCardz also claims that Nvidia’s board partners don’t expect this to change anytime soon.

QbitLeaks

Let’s go back to the configurable TGP. According to VideoCardz citing its own anonymous sources, the RTX 4090 will indeed have a default TGP of 450 watts, but users can adjust that in the BIOS to up to 660 watts. In the case of the RTX 4080 16GB, the base TGP (340 watts) can be increased to 516 watts, and for the RTX 4080 12GB, the standard 285 watts can go up to 366 watts.

This is an interesting, though unconfirmed feature. It would allow users to choose between an even higher power consumption and a performance boost. It looks like if it’s released it might not make it to every version of the map.

Soon most of this will be confirmed or debunked. Nvidia would drop the big reveal on September 20 as part of the GTC keynote.

Leave a Comment