With a week to go now until we get the much-wanted official unveiling of Nvidia's new RTX 4000-series graphics cards, we've just had leaked specs for the range's flagship GPU, the RTX 4090, as well as two variants of the incoming RTX 4080.
And, for PC gamers looking to upgrade to one of these new best graphics cards, the leaked specs reveal something very interesting about the RTX 4000-series of cards, and a very big reason for hardcore gamers to plump for the range's flagship RTX 4090 GPU.
This is because, as reported by ExtremeTech (opens in new tab), the specs gap between the top-level RTX 4080 and the RTX 4090 graphics card is noticeably bigger than it is in the current RTX 3000-series.
The Ampere architecture-based RTX 3000-series has a rough 10 per cent power difference between the RTX 3080 and 3090, but looking at the leaked Ada Lovelace architecture RTX 4000-series card specs, that performance gap seems to be a lot bigger.
According to the leaked specs, the RTX 4090 will come with 24GB of GDDR6X memory, boast 16,384 CUDA cores, have a 384-bit memory bus, a maximum bandwidth of 1,008 GB/s and a maximum TGP of 660W.
Meanwhile, the top-spec RTX 4080 will offer 16GB of GDDR6X memory, deliver 9,728 CUDA cores, have a 256-bit memory bus, a maximum bandwidth of 736 GB/s and a maximum TGP of 516W.
On paper, those specs say that the RTX 4090 GPU will offer much more than a 10 per cent power increase over the RTX 4080, and a very enticing reason for PC gamers to plump for the new flagship.
The T3 take: a proper flagship GPU
I think it fair to say that it makes sense for Nvidia to make the gap between its flagship new GPU and its other cards bigger. I mean, after all, you want your top product to be standout in every way, to be aspirational and a thing of awe.
But if your flagship isn't markedly better than the graphics card below it then you don't give PC gamers a very good reason to upgrade and spend the, no doubt, large extra premium to get the best GPU on the market. So, from a RTX 4090 point of view I'm on board Nvidia's train of thought here for sure.
Where things get a bit more cloudy is in the concurrent release of two variants of the RTX 4080, a 16GB and 12GB variant. These clearly will offer different levels of performance out of the box, but how they stack up against RTX 3000-series cards in terms of performance will be key, as too their price points.
As while the RTX 4090 was always going to command a top, top-tier price point, with the RTX 4080 cards (and especially the 12GB variant) now markedly far away from the flagship in terms of specs and power, how attractive they will be to gamers looking to upgrade remains unclear. Will PC gamers be better pluming for the Nvidia RTX 4060 instead, for example, and committing to the mid-tier fully.
The new Ada Lovelace architecture will obviously offer efficiency and feature improvements that are separate to raw power, but will they be enough to tempt PC gamers away from RTX 3000-series cards if the raw power increase isn't that marked?
Only time will tell. Be sure to check back in to T3.com soon for the official full picture on Nvidia's new RTX 4000-series graphics cards, which are scheduled to launch next week.