FinalWire announced a new version of its AIDA64 benchmark and diagnostic tool, and among the highlights is preliminary support for the NVIDIA GeForce RTX 4090, the green company’s upcoming top-of-the-line GPU. Unfortunately, the Beta update to reach version 6.70.6033 does not reveal any other details beyond confirming the model name and that it will be based on the NVIDIA AD102 graphics chip.
This serves to know that the first RTX 40 Series GPUs would already be tested, so it is logical to expect that in a matter of days, or at most weeks, we will begin to see performance leaks.
From what we know so far, the NVIDIA GeForce RTX 4090 will specifically use the GA102-300 graphics chip, which would hide no less than 16384 CUDA Cores inside, the frequencies that the chip could reach being a secret, for now, although there are Rumors that it would move around 2.75 GHz, nothing unreasonable if we take into account that its main rival, AMD, will use the same 5nm manufacturing process as TSMC, and it is rumoured that it will reach frequencies close to 3.00 GHz.
The graphics chip will be accompanied by 24 GB of GDDR6X @ 21 Gbps memory along with a 384-bit memory interface, while the TDP would be 450W, although there will be Custom models that significantly increase the TDP with configurations of two 16-pin PCI-Express 5.0 connectors. Due to this, we will have even more oversized cooling systems, while it is rumoured that the reference graphics, the Founders Edition, will incorporate a third fan to deal with the increase in temperatures.
Specifications of the NVIDIA GeForce RTX 4080 and GeForce RTX 4070
Along with the RTX 4090, we also know the specifications of these two models, with the RTX 4060 being the only one that has not yet succumbed to leaks, perhaps because this model has always been listed for early 2023.
The NVIDIA GeForce RTX 4080 would arrive with the AD103-300-A1 graphics chip with 10240 CUDA Cores (256 less vs RTX 3090) along with 16 GB of GDDR6 memory.