NVIDIA GeForce RTX 4060 Rumored – AD107, 115W Only

NVIDIA GeForce RTX 4060 Rumored – AD107, 115W Only

Significantly lower TDP compared to its direct predecessor.

Kopite7kimi is back again with another information drop – this time the RTX 4060. It is set to use AD107 die, which is the first time in xx60 SKUs. For reference, RTX 3060’s die is GA106 (for those who didn’t know – the smaller the third number is, the bigger and more powerful the full die is).

RTX 4060 still uses PG190.
AD107-400-A1
3072FP32
8G GDDR6 18Gbps
115W
24M L2

I will try to remain neutral about any leaks in the future. 😁😁😁

— kopite7kimi (@kopite7kimi) February 13, 2023

Looking through the specs you may noticed a few interesting things. Firstly, the CUDA cores are, in fact, less than its predecessor at just 3072 of them; VRAM gets downgraded to 8GB 128-bit 18Gbps dies which accounts for 288GB/s maximum bandwidth. In turn, the TDP gets cut one-third from 170W to just 115W, and that meant it’s technically possible that the chip may be carried to laptops, with all the specs intact – rare for mobile GPUs these days.

For easy reference, below is the comparison table of both xx60 cards:

GPU	RTX 4060	RTX 3060
Graphics die	AD107	GA106
CUDAs	3072	3584
VRAM	8GB GDDR6 128-bit, 288GB/s	12GB GDDR6 192-bit, 360GB/s
L2 Cache	24MB	3MB
TDP/TGP	115W	170W

There’s no word on the release date just yet, but if anything, it’ll be interesting to see the performance difference between the two when RTX 4060 gets out in the wild.

Source: Videocardz

Pokdepinion: The specs doesn’t look that promising to me… I hope NVIDIA isn’t going to resort to DLSS 3 just to say it’s faster than RTX 3060 though.