I also watched that video and I can tell you his numbers aren’t the final representation of the 7900xtx performance, I watched 4 other review videos for this card and in 3 of those 4 MSFS was benchmarked and it did much better than in Paul’s system, I’m not sure what the specs were though. There’s also a Verge review where msfs was used for comparison and the 7900xtx beats the 4080.
Now even there the results are actually similar: better at 1440, loosing at 4k. Also noticeable, this story about coil whine…
I wanted to give AMD a shot, i really did. But i need something reliable, and here it is more about a shot in the dark, praying for drivers update, as well as better watt management and fan control, which starts to make the list long…
Just keep in mind these GPUs are in their early stages.
I have had little to no issues with my rx 6800 in msfs. Much better performance and gaming stability all over my Nvidia cards.
Part of me believe this sim is better optmised for AMD hardware at least with DX12. Zero issues with DX12 except the texture pop outs.
Nvidia has the upper hand in 4k and better in VR. For 1080p,1440p AMD is the better option.
If you decide to stick with Nvidia and get the 4080 or 4090 you can use TAA with frame generation. DLSS is blurry on the instruments but it is better than FSR but neither are better than TAA ATM.
We should have a clearer picture in a couple of months when board partners release their cards and hopefully better drivers get released.
NVidia has become the scalper in a way, so I don’t want to support them this generation. So looking to see how the 7900XTX will evolve (I have a 3080Ti right now and an upgrade isn’t justified with those numbers, unless the upgrade is the 4090)
I think it is important to note that at lower resolution, you are CPU bottlenecked and it makes less of a difference. Also this was tested with DX11, for AMD I think DX12 runs slightly better on my 6800XT.
Ok, so this is the type of video you should steer clear of - It was released before the embargo and is showing numbers that do not align at all with other legit reviewers out there. The whole list is here, if you want to dig in: https://videocardz.com/144834/amd-radeon-rx-7900-xtx-xt-review-roundup
To cut short, the one that specifically look at MSFS are the following ones:
As far as I’m concerned, the only MSFS test that matters is with DLSS 3.0 Frame Generation turned on (and equivalent AMD feature for comparison), and in this test the result is:
Do you want the card that gets 95 fps over NYC in MSFS (avg 82), or the one that gets 132 fps (avg 126), at a lower power draw and with quieter fans? I know which one I’m leaning towards.
I have nothing good to say about “frame generation” and quite frankly will most likely never use it as I always lean towards quality and fidelity over the latest set of smoke and mirrors.
To me the bottom line is expressly without such such and if I’m consistently over 60fps everything else is sauce turn up the graphics to moarpleez. Also finance, power draw and home safety come to mind.
Also the 4080 should have been badged 4070 or even 4060 it’s so comparatively crippled for the cost.
Have you played on a 4080/4090? I’ve asked 4080/4090 users and they seem to think Frame Generation works great. The SimHanger guy did a video on it and even replied to me saying that he does not see any issues with FG in MFSF.
Anything over 60fps is a total waste of processing power; I’ve been most concerned about the frame-generation experience with low base frame rates, such as when people are stuggling to reach as high as 25-30fps landing at a busy airport full of AI aircraft (and sometimes get much lower than that, with very spiky up/down rate changes).
Is it at least “not worse” as without frame generation? Hopefully. It does mean that input lag is twice what you’d think it is from the visible frame rate, but for the most part that shouldn’t matter.
No idea, I’d like to see a detailed comparison in MSFS specifically where they talk about this. This video above is the best I found so far, with FSR enabled.
Nvidia is better only with DLSS. But without DLSS it’s equal.
I don’t understand why that matters. You buy a 4080 to play with DLSS enabled. If you don’t want DLSS you don’t buy one. From what I’ve seen it’s a game-changing tech for a CPU-limited sim like MSFS, there’s no real reason not to use it. The only issue I’ve seen is possible artifacts in POI markers in MSFS.
Anything over 60fps is a total waste of processing power
Do you play with a VRR display? I can absolutely tell the difference. Also getting much higher than 60 fps means that the lowest framerates will still be high(er), and it gives you headroom for future more complex scenery and higher detail levels.
Yes, but if I can tell the difference (I’d have to put them next to each other) I don’t care. It makes no difference to my experience, which is also mostly a still image with things moving slowly around.
Again, I don’t care and never will if it can reach 120 or 144 or 240 fps with no load at high altitude in the desert. Zero utility to that.
Takeoff and landing at busy airports. Measure that, and compare that.
Takeoff and landing at busy airports. Measure that, and compare that.
Agree, definitely the most important measure, I really want to see a comparison test of these two cards at busy airports like Heathrow. With AIG/FSLTL traffic.
Hopefully. It does mean that input lag is twice what you’d think it is from the visible frame rate, but for the most part that shouldn’t matter.
