Nvidia’s RTX 4080 graphics card could be a little less powerful than previously rumored, which quickly follows news that the RTX 4070 could be a spicier GPU than expected.
This new Lovelace speculation comes again from Kopite7kimi, a regular hardware leaker on Twitter that has been spreading quite a few rumors lately (though that’s to be expected as the RTX 4000 series approaches its launch time, with an expected debut in or around October). , assuming there are no delays).
I’m not a chatterbox, but I need to make some updates. I hope you do not mind. a possible RTX 4080, PG136/139-SKU360AD103-300-A19728FP32256bit 16G 21Gbps GDDR6Xtotal power ~420WTSE ~15000Now I have completed the latest update for 4090, 4080 and 4070.August 5, 2022
The theory is that the ‘possible’ RTX 4080 – a reminder that nothing is set in stone per se, even if this is Nvidia’s current thinking with the GPU’s specs – could run with 9,728 CUDA cores, which is less. than the previously rumored 10,240 .
The leaker believes that the RTX 4080 will be equipped with 16 GB GDDR6X (at 21 Gbps) and a power consumption of around 420 W, all of which remains in line with what was previously suggested – the only change is that reducing the number of cores at this point.
Analysis: Nvidia’s positioning of RTX 4000 models
The first thing to note here is that this is not a big change. Reducing 512 CUDA cores — meaning the onboard streaming multiprocessors are cut by four, from 80 to 76 — isn’t a huge move, but it still means a slight reduction in raw grunt for the GPU.
Let’s keep in mind that this is just a rumor, and the leak landscape around Lovelace seems to be shifting quite regularly these days – which could be a reflection of Nvidia tinkering with and tweaking relative specs as the company moves towards a definitive number of products.
As mentioned at the outset, Kopite7kimi also recently brought us news about the RTX 4070, with the ostensible idea that at this stage that GPU would boost its level of performance – to the specification previously rumored for the RTX 4070 Ti, in fact. Quite a step up, that’s for sure. So, assuming this is all correct, what could Nvidia’s reasoning be here?
Decreasing the RTX 4080’s CUDA Core count a bit, while seriously boosting the RTX 4070 – not only for CUDA Cores, but also for the load on VRAM – clearly means that the relative performance of the RTX 4070 and RTX 4080 is theoretically more aligned.
With less differentiation between the two, and Nvidia’s xx80 model is generally quite a bit more expensive than the more mainstream xx70 card – the relatively affordable high-end GPU, if you will – with Lovelace, we’re going to be looking at an RTX 4070 what else is in the sweet spot for most buyers?
Or to put it another way, if the RTX 4080 isn’t all so much faster, who goes for one instead of a 4070? Unless of course stock is an issue for the latter; that could play a role, especially if it’s really popular.
We should also remember that CUDA cores are not the full story for performance, with many other factors at play, such as clock speeds. It’s here that there could be an indication that the RTX 4070 is supposedly sticking to around 300W for power consumption, while the RTX 4080 will reportedly hit 420W or thereabouts – this could point to Nvidia seriously increasing clock speeds (and thus performance) with the more expensive GPU compared to the 4070.
Furthermore, Kopite7kimi has kept the previous guestimate for a 3DMark TimeSpy Extreme score for the RTX 4080 (about 15,000), despite those CUDA cores being shaved off, so interestingly enough, that could also suggest that Nvidia might be pushing harder with clocks for this. the 4080 Although the leaker does not specifically mention clock speeds.
Surely it makes no sense to have the RTX 4080 as a relatively unappealing proposition compared to the 4070 – much higher power consumption won’t help either, in terms of forcing PSU upgrades and power costs in general – so for now, we’re assuming that if this rumor is true, we can’t yet see the full picture of what Nvidia is trying to do with the positioning of Lovelace GPUs. Or that there will be yet another update of the specifications …
Relative pricing could also be a factor here – we haven’t heard much from the rumor mill on that front, but we can’t imagine launching the RTX 4000 series with affordability in mind. Especially not when Nvidia has a lot of excess RTX 3000 stock to deal with, as the grapevine claims, and shifts anything that overlaps with Lovelace’s launch (which is probably only a few months away now).
Be that as it may, you could imagine that Team Green would have learned from its mistakes in terms of the RTX 3080 getting off to a bad start, with only 10GB of VRAM in its first incarnation – an unpopular move that has yet to become exacerbated by the later release of a 12GB variant. The latter wasn’t just more powerful in terms of the extra video memory, which made some buyers of the original RTX 3080 deeply remorseful.
Via VideoCardz (opens in new tab)