Today, NVIDIA is announcing three brand-new graphics cards launching this November with prices starting at $899. There are two RTX 4080 variants, one 12GB and one 16GB, while the RTX 4090 comes in with 24GB of memory. As NVIDIA’s latest graphics cards, there’s a lot to look at to see whether it’s worth picking up the new generation or sticking with sales on the 30-series. So, without further ado, let’s dive into NVIDIA’s RTX 40-series launch.
NVIDIA launches RTX 40-series graphics cards with November release date
We’ve all been waiting for the announcement, at least, if you’re into PC gaming at all. Well, it’s finally here. The RTX 40-series from NVIDIA is officially official and launches in November with a starting price of $899.
The tagline for this launch is “Beyond Fast,” and NVIDIA is really pushing its new Ada Lovelace architecture here. Interestingly enough, there’s no RTX 4070 announced yet (though it could still come), but we did get two RTX 4080 models with varying memory options. The base RTX 4080 comes with 12GB of GDDR6X memory, 7680 CUDA cores, and 192-bit memory bus. The rest of the specs are the same with the higher-end 16GB RTX 4080, though it does up the CUDA core count to 9728 and the memory bus to 256-bit.
This means that there’s more than just a VRAM difference here, and that the 4080 16GB will usher in additional performance across the board. These differences aren’t small, either. While they both have 3rd Generation ray tracing cores, 4th Generation Tensor cores, DLSS 3, and all the other features, the bump in CUDA cores and memory bus will give improved gameplay performance across the board, for sure. We saw that a bump in memory bus and CUDA cores allowed the RTX 3090 Ti to eek out a little more performance over the 3090, and that was with the same amount of VRAM.
Moving onto the RTX 4090, the memory steps up to 24GB of GDDR6X on a 384-bit memory bus. That’s the same amount of memory and same bandwidth as the RTX 3090 Ti, interestingly enough. However, the CUDA cores are upped to 16384 from 10752, which is a fairly substantial increase. Also, the 4090 keeps the max power at 450W, which is also on par with the 3090 Ti.
NVIDIA’s latest RTX 4080 12GB will cost $899, the 16GB model will run $1,199, and the RTX 4090 has a price tag of $1,599. The RTX 4090 will be available to purchase October 12 and the 4080 will be available in November.
I’m both pleased and a little surprised with NVIDIA’s launch today. It looks like there will be some good improvements across the board with RTX 40-series, but the fact that all cards are still on PCIe 4.0 and none have made the jump to PCIe 5.0 leaves room for improvement on AMD’s side when they ready their next-generation cards. There’s a lot more bandwidth available to PCIe 5.0, just take a look at the SSD theoretical maximum speed increase from 7GB/s to nearly 13GB/s. So, why NVIDIA didn’t ready RTX 40-series for PCIe 5.0 with backward compatibility with 4.0 systems is beyond me.
Also, I fully believe the 12GB RTX 4080 should be the RTX 4070. It has less memory, fewer CUDA cores, and a smaller memory bus. It’s a weaker card all around, not just in the VRAM department. Why NVIDIA chose to confuse gamers and call it the same name as the 16GB variant I have no idea, because the 12GB GPU even has a lower 285W TDP verses the 320W TDP of the 16GB model. It’s odd all around, honestly, but only time will tell if the difference is as great in real life as it is on paper.
I’m excited to see what NVIDIA’s RTX 40-series has to offer, as the 30-series ushered in some crazy performance gains over 20-series across the board. Will the 40-series do the same from NVIDIA?
FTC: 9to5Toys is reader supported, we may earn income on affiliate links