Not everyone runs at 4k mate and many likely never will. 4k currently makes up 2.25% of all PC's that took part in the Steam Hardware Survey. 4k is not and likely never will be the benchmark for gaming performance. VRAM is like a gas in a sealed room. Games will take advantage of it even when it provides no real performance boost.
I think people need to stop quoting Steam hardware surveys, when it includes every low-end rig from every 2nd and 3rd world country that has Steam and an internet connection.
That figure of 2.25% (which is actually still significant if you think how many PC's it polls) would look very different if it was further filtered by geography. People in the richer first-world countries (for example Europe / USA) clearly buy proportionally more high-end PC hardware than otherwise and that it where the majority of high-end PC hardware sales go and where playing at 4k is gaining far more traction.
Well no there’s more than one, MSFS 2020, Control, Watchdogs, RDR2 (when maxed) and of course we are right at the start of the 3000 series now and games are only going to get more intensive. I do agree with you though that the 3080ti is only a marginal upgrade over the 3080 whereas previously the ti model has been a good 30-40% better rather than 5-15%.
- FS2020 is not intended for high FPS... the engine grinds anything it gets and it is also very CPU dependent. It doesn't even use DX12... this is clearly a case where the development is affecting performance and it is well documented online.
- Watchdog Legion is unoptimised crap (it's Ubisoft and they are and always will be developer dirt). Another case where the development is affecting performance.
- RDR2 runs at or around (sometimes a bit higher sometimes a bit lower) 60fps at 4k maxed out... for a slower-paced game that looks as good as it does, with that insane amount of world detail, that is perfectly acceptable. Rockstar knocked it out of the park with this game and it is just stunning even 2 years later.
The vast majority of titles can be played at 60fps or more at 4k (usually without RT, or enabling DLSS with it), this is backed up by the review benchmark suites you can find on Youtube. There will always be, and always have been, exceptions that perform worse on new cards for a variety of reasons. Game development and architecture is not consistent and one rule does not fit all.
Hang on you said it is a marginal upgrade (which I agree with) and there's no point in spending extra money on fancy AIB versions? Until you know which 3080Ti Dave ends up with I would keep the sock in the mouth as 1) he has to obtain one (as we know they will be in short supply), then 2) sources one of the entry priced cards to escape your assumption?
This paragraph makes no logical sense.
Although I will
learn to read in context,
there is nothing smart about pointing out any custom clocked AIB marketed flavour of the 30 series lineup tend to be way overpriced when compared to the FE. The whole sales pitch by Jensen was a rock bottom MSRP such as the £649 FE of the 3080 yet we know in reality anyone who didn't get them will have paid likely more for any AIB flavour. This holds true for all of them. The MSI gaming X Trio would unlikely have been £649.
Again Th0nt... I am struggling to see how you don't get this. There are many different level of AIB cards and the ROG is one of the worst for being overpriced. I clearly mentioned I was talking about the "no frills" AIB cards... those are always closer to MRSP levels. Dave wrote that he wants to pick up an FE, or at least one close to MSRP and
he is willing to wait... so lets use our grey matter and assume that he will be waiting for one at the lower-end of the pricing scale.
A 50% extra cost on the baseline 3080 makes it £973. Im not really sure where your getting your 15% performance increase from but I will wait and see now how accurate everyone is once they appear.
Again, this is not rocket science, to the point where I am a little confused as to how you can even write that. The 3090 is on a 'ballpark average' 10-15% faster than the 3080 (despite costing an unjustifiable 100% more, ugh) and the general assumption is that the 3080Ti is going to perform at or around (ie: close to) 3090 levels. Except, it should instead cost around 50% more.
I can't be any clearer than that so if you still don't understand then lets simply agree to disagree as I don't want to waste any more time explaining basic things that
should be clear to anyone with a bit of common sense.
