• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

30xx Series Founders Edition

Do you have the Reddit channel? I’m toying between 4090 and 3090ti and play MsFS


Another one seeing the same thing but he updated from a 3090ti to 4090, no difference on a monitor but VR gets a good boost.
:cry:


Also watch for the tearing in VR (the side window area tearing like crazy every time he moves the view) and 3090ti beats the 4090 with DLSS on with a monitor..
:eek:


So I wasn't going crazy when I tested 3090 vs 4090 on a 7950x.

Another one :-



EditerTech Alpha Insider
2d

I didn’t do thorough recorded benchmarks or anything, but I did do some baseline performance comparisons to see the performance difference between the RTX 3080 10GB and the RTX 4090 24GB on my existing system, with a 10900K, 32GB RAM.
  • In heavily CPU-bound situations (settings on Ultra, poor performance areas like NYC/London/Seattle), the sim is completely CPU bound even under DX12, and my frame rates were basically identical (within 1-2 fps) without the DLSS Frame Generation turned on. Once I turned that on, FPS mostly doubled. From 40ish around NYC to about 75 on average in the same area/plane/weather. (5,120x1,440 monitor, so essentially similar rendering demands as a 4K display.)
  • There’s a been a lot of talk about visual glitches with frame generation turned on. I have never seen/noticed such a glitch. I could probably find some if I video-captured a flight and frame-by-framed, but I’m not seeing/noticing them in actual flying.
  • In VR, though, I did see a pretty big increase in frame rate over the 3080, even without DLSS 3.0/Frame Generation being available. Because I’m less CPU-bound there (I had the LOD turned down to 100 to maintain a good frame rate on the 3080), I went from around 45 fps to about 70-72 on a Reverb G2, which is dramatically faster. This is with motion reprojection off. (MR artifacts bug me.) Enabling DLSS quality gave me about the same frame rate as 80% resolution scaling but a little sharper image.
No regrets on the upgrade, but in my limited testing of a couple of hours of playing around with the card, my initial impressions are:
  • In games, which are usually GPU-limited, the 4090 is a beast. In MSFS, which is easy to CPU-limit, it’s still a beast, but only because of the frame doubling. If you’re CPU-limited, you may not see much difference over a 3080 unless you run DX12 with frame generation enabled.
  • Given that all the real increase is coming from the frame doubling, rather than rushing to get a 4090, watch to see what kinds of frame rates people get in MSFS with 4080s or even 4070-level cards. If you’re seeing your base performance CPU-limited, those cards will also offer DLSS Frame Generation, and you may get identical performance at a lower price.
  • The moral of the story is you’re gonna need both a killer CPU and a 40-series video card to get the most out of MSFS.
I have a 13900K on order, so it’ll be interesting to see how much that combo increases performance.


Basically zero difference on monitor users and only a benefit for VR and DLSS 3.0 doesn't work for VR DLSS2 does and that shows the uplift at VR resolutions but people complain the image is blurry too, because frame generation can't work on VR and never will frame generation will make a mess and make it unusable. Meeh ... Nvidia pulled a scam really with this as a selling point and DLSS 3.0 is not how people want to use MSFS in DX12 as DX12 is broken in the game and still in BETA, anyways more testing I need to do this weekend so far not seeing any changes from my 3090 to a 4090 at 4k with a 7950x at native DX11 and DX12 same FPS just less GPU utilisation on the 4090 ..because still cpu bound.


Another one mentioned before with a 3090 to 4090 :-

 

Another one seeing the same thing but he updated from a 3090ti to 4090, no difference on a monitor but VR gets a good boost.
:cry:


Also watch for the tearing in VR (the side window area tearing like crazy every time he moves the view) and 3090ti beats the 4090 with DLSS on with a monitor..
:eek:


So I wasn't going crazy when I tested 3090 vs 4090 on a 7950x.

Another one :-






Basically zero difference on monitor users and only a benefit for VR and DLSS 3.0 doesn't work for VR DLSS2 does and that shows the uplift at VR resolutions but people complain the image is blurry too, because frame generation can't work on VR and never will frame generation will make a mess and make it unusable. Meeh ... Nvidia pulled a scam really with this as a selling point and DLSS 3.0 is not how people want to use MSFS in DX12 as DX12 is broken in the game and still in BETA, anyways more testing I need to do this weekend so far not seeing any changes from my 3090 to a 4090 at 4k with a 7950x at native DX11 and DX12 same FPS just less GPU utilisation on the 4090 ..because still cpu bound.


Another one mentioned before with a 3090 to 4090 :-

Really interesting results. I’ve no interest in VR and I have a similar 32:9 monitor setup. Im not too fussed by DLSS as I like my quality non blurry cockpit! Looks like I’ll settle on a 3090ti, save a few quid in the process and this will see me through until the 4090ti and beyond.
 
10% faster in a synthetic benchmark so basically no quicker at all in real games.

The 3060Ti isn't exactly bandwidth starved in the same way the 3070 wasn't.

Also stating the new card is quicker than the A750 in 3Dmark is not really a massive win. I own a A750 and whilst it is faster than my 3060Ti FE in a number of synthetics it is currently getting pummelled in actual games. This may change in the future (although I doubt a role reversal) that is the reality currently.
 
10% faster in a synthetic benchmark so basically no quicker at all in real games.

The 3060Ti isn't exactly bandwidth starved in the same way the 3070 wasn't.

Also stating the new card is quicker than the A750 in 3Dmark is not really a massive win. I own a A750 and whilst it is faster than my 3060Ti FE in a number of synthetics it is currently getting pummelled in actual games. This may change in the future (although I doubt a role reversal) that is the reality currently.

Still funny to see this from a company who just done an embarrassing amount of back pedalling over their rebadged 4070 fiasco.


Not to mention 3080FE and then 3080FE LHR. And you can also chuck the 12 gig 3080 in there even though there was no FE model of that. But after all that our little brains can't process 4080 12 gig and 4080 16 gig. :cry:
 
Still funny to see this from a company who just done an embarrassing amount of back pedalling over their rebadged 4070 fiasco.


Not to mention 3080FE and then 3080FE LHR. And you can also chuck the 12 gig 3080 in there even though there was no FE model of that. But after all that our little brains can't process 4080 12 gig and 4080 16 gig. :cry:

Clown era. Still 80%+ lapping them up though. Hopefully be 75% after this gen..
 
Really interesting results. I’ve no interest in VR and I have a similar 32:9 monitor setup. Im not too fussed by DLSS as I like my quality non blurry cockpit! Looks like I’ll settle on a 3090ti, save a few quid in the process and this will see me through until the 4090ti and beyond.

Yep, good feedback @Purgatory . Need them real world instances and not some company bar charts trying to sell their products to base performance on. 2X is suddenly now 65% in wider tests.
 
Yep, good feedback @Purgatory . Need them real world instances and not some company bar charts trying to sell their products to base performance on. 2X is suddenly now 65% in wider tests.

Those synthetic benchmarks are generally well accepted as reasonable indicators of performance. Not everyone can code games like Id software
 
Last edited:
The 3060Ti was an epic card tbh, I had one for a while before I got my 3080, price performance of both were spot on tbf but if this ones a bit faster than the 'old' 3060Ti then whats happening to the 3070?
Maybe Nvidia are killing off the x070 cards lol
 
Still funny to see this from a company who just done an embarrassing amount of back pedalling over their rebadged 4070 fiasco.


Not to mention 3080FE and then 3080FE LHR. And you can also chuck the 12 gig 3080 in there even though there was no FE model of that. But after all that our little brains can't process 4080 12 gig and 4080 16 gig. :cry:

Oh yes sorry I wasn't disagreeing with your point (the whole 4080 12GB and "oh sorry we thought it was a little confusing" combined with the tech tubers basically giving them a free pass is an utter joke).

Just thought it was amusing where they were using the A750 as a comparison. It really should not be used in this manner.

I will say though on the 3060Ti that it is absolutely perfect for my use case, 1440p high refresh and high (not pointless max) settings. Doing some testing in Destiny 2 and it can hold over 120FPS in the more demanding throne world settings which is quite impressive. Yes my "old" 3090 would show it the door but anything north of 144Hz (at 1440p UW in its case) was a complete waste. It still crumbled in CP2077 with RT anyway so the extra grunt was relatively pointless for me.
 
So bloody frustrating, really would love a 3080 to replace the 2070 super I have, and then put the 2070 in my other 980 GTX system. But Car MOT must come first! The urge to spend is really bad lol

3080's should be around £400 now, 500+ is still way too much
 
So bloody frustrating, really would love a 3080 to replace the 2070 super I have, and then put the 2070 in my other 980 GTX system. But Car MOT must come first! The urge to spend is really bad lol

3080's should be around £400 now, 500+ is still way too much

They'll stay where they are until stock dries up, then after the 30 series is all sold out we might see a £50/75 drop, IMHO..
 
Back
Top Bottom