RTX 4080 & 7900 XTX/XT benchmarks DCS

[xoxen]

This is my first video and regarding the settings I followed the ones mentioned before:

For me it looks like the 4080 is slower than the 7900XTX. I opened the track file and ran it. Comparing different scenes show sometimes more, but overall less FPS. Not sure how to judge this for now. Actually my CPU is undervolted, but it runs with the same 4850 MHz than in the other system.

Edit: I carefully compared the videos. It is not a clear picture. The 7900 starts strong, has a weaker phase but is stronger when flying over the river later on. There the 4080 dips more down, even into the 50s. My conclusion is that both have it strengths. Hard to say.

Edit 2: For the video I used the Nvidia driver 526.98 WHQL, I just installed the newer version 527.56 WHQL and it performs better. I will create a video tomorrow as I´m tired and I´m ready to go to bed.

Edited December 16, 2022 by xoxen
Additional information

[frachy]

2小时前，Th3ChosenOn3说：

Seems I won't be able to test a resolution above 3440x1440. Unfortunately I can't seem to get the game to behave properly with a resolution of 3840*2160 on my 21:9 aspect ratio monitor.

Sorry.

Nevermind, Thank you : )

[dock999]

Any chance to test the 4080 and 7900XTX in DCS VR environment?  

[xoxen]

6 hours ago, dock999 said:

Any chance to test the 4080 and 7900XTX in DCS VR environment?  

I don't have this chance, sorry.

[Lange_666]

10 hours ago, xoxen said:

Edit 2: For the video I used the Nvidia driver 526.98 WHQL, I just installed the newer version 527.56 WHQL and it performs better. I will create a video tomorrow as I´m tired and I´m ready to go to bed.

Any chance you can put the difference between the two in here?

[Andrei]

Really good effort, thanks guys! I think those benchmarks you did are really valuable as it's one of those rare occasions that allow us to compare more or less apples-to-apples.

The results are interesting indeed. I'd say both cards are pretty close overall. Perhaps 7900 XTX works slightly better with CPU thus the better frame rate when flying over the river which is quite action packed.

VR will not be possible to test it seems. I'd care a guess based on prior experience that 4080 will give somewhat more consistent frame times, but it's hard to tell for sure. To me then both cards seem to be a decent option ignoring the price  

I will probably go for the one I can get based on the availability / price (discounts) and board design in terms of cooling and noise.

[xoxen]

1 hour ago, Lange_666 said:

Any chance you can put the difference between the two in here?

Here you go:

It takes some time until the HD version is up, it was mentioned 45 min.

[frachy]

12分钟前，Andrei说：

Really good effort, thanks guys! I think those benchmarks you did are really valuable as it's one of those rare occasions that allow us to compare more or less apples-to-apples.

The results are interesting indeed. I'd say both cards are pretty close overall. Perhaps 7900 XTX works slightly better with CPU thus the better frame rate when flying over the river which is quite action packed.

VR will not be possible to test it seems. I'd care a guess based on prior experience that 4080 will give somewhat more consistent frame times, but it's hard to tell for sure. To me then both cards seem to be a decent option ignoring the price  

I will probably go for the one I can get based on the availability / price (discounts) and board design in terms of cooling and noise.

yes i guess the same, these two cards are pretty close, maybe 7900XTX is slightly better in fps, not much, but it has 100W idle power...

[Th3ChosenOn3]

17 hours ago, xoxen said:

This is my first video and regarding the settings I followed the ones mentioned before:

For me it looks like the 4080 is slower than the 7900XTX. I opened the track file and ran it. Comparing different scenes show sometimes more, but overall less FPS. Not sure how to judge this for now. Actually my CPU is undervolted, but it runs with the same 4850 MHz than in the other system.

Edit: I carefully compared the videos. It is not a clear picture. The 7900 starts strong, has a weaker phase but is stronger when flying over the river later on. There the 4080 dips more down, even into the 50s. My conclusion is that both have it strengths. Hard to say.

Edit 2: For the video I used the Nvidia driver 526.98 WHQL, I just installed the newer version 527.56 WHQL and it performs better. I will create a video tomorrow as I´m tired and I´m ready to go to bed.

 

So I've gone and taken your second run with the updated driver, and ran another test with the track file I sent you to allow for a frame by frame comparison. Considering that our system are slightly different in terms of motherboard, ram, storage, and the fact that the track will have slight variation between runs, it's important to try and average the results and not take the values at face value.

Looking at them side by side, I'd say they perform on par with one another. I believe the slightly higher FPS on the RX 7900 XTX is due to my ram/storage configuration being slightly faster than yours, especially since both our system's aren't maxing out the GPU's to 100%. There's a couple of frames where we are at the same FPS and both cards are showing the same GPU percentage. We might have to run another test where the GPU's are being maxed out, instead of being limited to our CPU's. It might also be worthwhile to come back to this test after ED's implementation of multithreading.

On the RX 7900 XTX side of things, there's glitchy smoke as seen in the clips. I'm not sure if it's a driver, game setting or issue with the game that may be causing it, but I've only noticed the effect in the benchmark. I also think the RX 7900 XTX uses quite a lot more power compared to the RTX 4080. My card will pull up to 400W on 100% usage in other games, and I don't think the RTX 4080 will pull that high, so may be something to consider if you're in a region with expensive electricity. My card is also pulling a consistent 126W while writing this post, but I believe it'll be fixed with a driver update in the future.

Also for VR, it's been mentioned before that the RX 7900 XTX will lose to the RTX 4080 at the moment, due to driver issues. I don't have a VR headset to test with, so I can't say at the moment how the card will perform and it'll be up to someone else to test that scenario.

For my conclusion, I'd say if you're not a VR gamer then both cards perform the same and you'd be happy with either but the deciding factor on which to get will depend on price, which card are in stock and if you're willing to deal with the AMD driver problems until they're sorted. 

 

Edited December 17, 2022 by Th3ChosenOn3
Added conclusion.

[SkateZilla]

7900XT/XTX Owners,

Use GPUz to Check your GPU Stepping, AMD Apparently Shipped Chips in the A0 Stepping that had hardware problems, if you got a Card with a A0 GPU, I'd advise contacting MFG for an Exchange.

[CybrSlydr]

 

[Vizharan]

vor 4 Stunden schrieb SkateZilla:

7900XT/XTX Owners,

Use GPUz to Check your GPU Stepping, AMD Apparently Shipped Chips in the A0 Stepping that had hardware problems, if you got a Card with a A0 GPU, I'd advise contacting MFG for an Exchange.

I've read somewhere that a AMD spokesman debunked that today.

 

Besides that as far as I've seen it every shipped 7900 XTX has C8 as Revision (mine too btw :D)

[AngleOff66]

AMD responds

https://www.tomshardware.com/news/amd-addresses-controversy-rdna-3-shader-pre-fetching-works-fine

[SkateZilla]

Jay is back from vacay..

Strongly recommended watch if you have a 7900XTX Reference Card. or any Card with Only 2x 8 Pin Connectors
 

 

Edited December 18, 2022 by SkateZilla

[BitMaster]

LoL !

Not even sure if that can all can be fixed with Software, like the idle draw issue.#

Oh man, none of those cards really shines, either you accept a potentially dangerous plug but overall great card or take the safer card, for less money too, but that has some quirks too we have no assurance if they can be and will be fixed by Software.

Please, 1080ti, last at least ONE more round  

[petsild]

On 12/17/2022 at 3:55 PM, Th3ChosenOn3 said:

So I've gone and taken your second run with the updated driver, and ran another test with the track file I sent you to allow for a frame by frame comparison. Considering that our system are slightly different in terms of motherboard, ram, storage, and the fact that the track will have slight variation between runs, it's important to try and average the results and not take the values at face value.

Looking at them side by side, I'd say they perform on par with one another. I believe the slightly higher FPS on the RX 7900 XTX is due to my ram/storage configuration being slightly faster than yours, especially since both our system's aren't maxing out the GPU's to 100%. There's a couple of frames where we are at the same FPS and both cards are showing the same GPU percentage. We might have to run another test where the GPU's are being maxed out, instead of being limited to our CPU's. It might also be worthwhile to come back to this test after ED's implementation of multithreading.

On the RX 7900 XTX side of things, there's glitchy smoke as seen in the clips. I'm not sure if it's a driver, game setting or issue with the game that may be causing it, but I've only noticed the effect in the benchmark. I also think the RX 7900 XTX uses quite a lot more power compared to the RTX 4080. My card will pull up to 400W on 100% usage in other games, and I don't think the RTX 4080 will pull that high, so may be something to consider if you're in a region with expensive electricity. My card is also pulling a consistent 126W while writing this post, but I believe it'll be fixed with a driver update in the future.

Also for VR, it's been mentioned before that the RX 7900 XTX will lose to the RTX 4080 at the moment, due to driver issues. I don't have a VR headset to test with, so I can't say at the moment how the card will perform and it'll be up to someone else to test that scenario.

For my conclusion, I'd say if you're not a VR gamer then both cards perform the same and you'd be happy with either but the deciding factor on which to get will depend on price, which card are in stock and if you're willing to deal with the AMD driver problems until they're sorted. 

 

 

Hi,

can you please increase the AA value to 4x and add SSAA at least x1.5 to rerun the benchmark, thank you very much.

[Johnny Dioxin]

Just ordered my Zotac 4080 along with a Pico 4 - should be interesting to see what performance they bring in flight sims. My current set-up has been good enough to run all sims smoothly in VR, but at half rates (30 or 45 fps - can only run FSX with FlyInside VR at full 120fps with Index)  and low-ish settings.

I was  hoping to upgrade my CPU, but can't be done sensibly - I have an i5 9600k and the only thing I can upgrade to with the Z390D mobo is the i7 9700k - just not worth it. I really don't want to buy a whole new rig just yet. Luckily, I've only had one sim that had my cpu maxed out, and it wasn't DCS World. That may change now - but I should still see a good jump in VR performance, regardless. I'll still have this card when I next upgrade the rest of my rig

Anyway, I'll post VR benchmark results with Index, Quest 2 and Pico 4 once I'm all set up. That'll give people with hardware a couple of years old who are thinking of upgrading their graphics card for VR an idea of whether it's worth it or not.

[MelkorBorgir]

think i might wait till a 7950xtx launch, my 6800xt is still doing well for me

2 hours ago, Johnny Dioxin said:

 

I was  hoping to upgrade my CPU, but can't be done sensibly - I have an i5 9600k and the only thing I can upgrade to with the Z390D mobo is the i7 9700k - just not worth it. I really don't want to buy a whole new rig just yet. Luckily, I've only had one sim that had my cpu maxed out, and it wasn't DCS World. That may change now - but I should still see a good jump in VR performance, regardless. I'll still have this card when I next upgrade the rest of my rig

can you not upgrade to 9900k?

 

[Johnny Dioxin]

9900k prices are still quite silly, so for the modest gain in performance it isn't worth the outlay. Last time I tried a cpu upgrade in the same family, I was very disappointed and ended up putting my original i5 back in. That was from an i5 4960k to the i7 4770k, which everyone was raving about. The overclocking was dire and even on standard speeds it gave no better performance than my i5. That's something I won't forget in a hurry. I know the i9 is another step, but it's just not worth the outlay right now, especially with me planning to upgrade to AMD in a year or so.

Not used AMD since my Athlon64 days - and they now run some of my retro PCs. That's providing nothing big changes in the next year or so.

edit: I also came across this on a testing website:

Quote

It's probably a waste to go from the 9600K to the 9900K already. The increase in performance would only be visible in some select games. For the most part, it would be so similar that you couldn't tell there's a faster CPU.

btw I know I said I'd wait until January for the 4080, but a good buy came along and I liked the look of the card and they only had one left - so I jumped. Cost me £1346 for a Zotac Carte Graphique Nvidia GeForce RTX 4080 Trinity OC 16Go

Edited December 19, 2022 by Johnny Dioxin

[ssg]

I finally got my 7900 XTX. I ran the Plazmas test also, but on F14. 
5800X3D, 32Gb RAM, SSD, 3080/7900XTX, Quest 2

The DCS settings are what I have settled for 3080 and Quest 2. I haven't played around with them yet. 

https://imgur.com/a/hWFhwYv

Pretty disappointing. I was expecting that VR performance is not good, but this card is not good on flatscreen either. I also tested Arma 3, there too this card performed worse than 3080. I suspeced that my card is not functionin properly, but I ran 3D Mark and I got above average score for this card so it is fine. 

The frametimes on graph seem more stable on 7900XTX, but in practice it was stuttery mess. Everytime you turn your head it just stutters. On 3080 there is no stutters. 
 

Edited December 19, 2022 by ssg

[shastasm]

Hi All.  So I did a little benchmarking on the 4080 this weekend.  Below are my results.  All my settings were set at high (mostly). Settings are posted below.  I ran with different SSAA and MSAA settings   This test was ran at 1440p and out of Batumi (Georgia Map) early morning with some clouds.  I tried to recreate the test each time using the F-16 by over flying the sea at 20 feet, then over some defensive units firing at me at the old cross-landing strip north of Batumi, then low level flight in the trees.  The biggest FPS impact in all cases was in the trees at low level.  Highest FPS was observed at high altitude in the 170 to 180 range.  I recorded all the test runs and debating whether or not to post on YouTube as its not a great video lol.  Below are my results.

 

I think I'll be using the MSAA at 4 - good balance with graphics and performance.  1% and 0.1% lows were better than the MSAA at 2.   

 

System:

Ryzen 9 5900x

64MB RAM - 3200Mhz

GTX 4080 16GB

 

Edited December 19, 2022 by shastasm

[VirusAM]

I finally got my 7900 XTX. I ran the Plazmas test also, but on F14. 
5800X3D, 32Gb RAM, SSD, 3080/7900XTX, Quest 2
The DCS settings are what I have settled for 3080 and Quest 2. I haven't played around with them yet. 
https://imgur.com/a/hWFhwYv

Pretty disappointing. I was expecting that VR performance is not good, but this card is not good on flatscreen either. I also tested Arma 3, there too this card performed worse than 3080. I suspeced that my card is not functionin properly, but I ran 3D Mark and I got above average score for this card so it is fine. 
The frametimes on graph seem more stable on 7900XTX, but in practice it was stuttery mess. Everytime you turn your head it just stutters. On 3080 there is no stutters. 
 

On amd cards reprojection doesn't work properly...i don't know about the new ones and probably drivers will also improve

[Dudikoff]

27 minutes ago, VirusAM said:

On amd cards reprojection doesn't work properly...i don't know about the new ones and probably drivers will also improve

Yeah, I'd presume it's due to the new chiplet design used on RDNA 3 and that the VR software will need updates to support that.

Edited December 19, 2022 by Dudikoff

[Th3ChosenOn3]

Hi guys, @xoxen and I ran another benchmark to compare the RTX 4080 against the RX 7900 XTX on our similar systems. This time on the Marianas map to put our machines in a more GPU bottlenecked scenario.

Xoxen kindly flew the mission, provided the track file and a recording of his system playing the track file to put both system in the same scenario for comparison. I've attached the track file for anybody else that would like to run the same benchmark/track to compare their system.

Overall, the RX 7900 XTX seems to be around 10-20% faster across the entire track. I've also gone and done a mild overclock on my card and ran the same track file for comparison, which resulted in another 10% gain over the XTX and around 30% at times faster than the RTX 4080. We used the same settings as before, which is Plazma's 1440p settings.

Marianas Benchmark.trk

Edited December 19, 2022 by Th3ChosenOn3
Spelling

[shastasm]

1 hour ago, Th3ChosenOn3 said:

Hi guys, @xoxen and I ran another benchmark to compare the RTX 4080 against the RX 7900 XTX on our similar systems. This time on the Marianas map to put our machines in a more GPU bottlenecked scenario.

Xoxen kindly flew the mission, provided the track file and a recording of his system playing the track file to put both system in the same scenario for comparison. I've attached the track file for anybody else that would like to run the same benchmark/track to compare their system.

Overall, the RX 7900 XTX seems to be around 10-20% faster across the entire track. I've also gone and done a mild overclock on my card and ran the same track file for comparison, which resulted in another 10% gain over the XTX and around 30% at times faster than the RTX 4080. We used the same settings as before, which is Plazma's 1440p settings.

Marianas Benchmark.trk 70.72 kB · 4 downloads

 

Thanks for sharing.  I just ran the track with the same settings and got about 5 to 10 more FPS than you with the 4080.  Could be a function of the different models (Asus TUFF for me).  Super glad there are options out there for our DCS friends.  I might have to overclock now to compare :).  Again thanks for sharing.