Jump to content

RTX 4090 benchmarks! if anyone finds any, please post here!


Recommended Posts

1 hour ago, BuzzU said:

Yes, but it's just a chart. I can see his running in DCS.

Thing is not the 15 FPS. It the frametimes that counts. Seeing this example, the 4090 might be the first card where you can run a stable 45fps reprojection with all settings cranked up to "high". You wouldn't even benefit from a 6090 because you would always be stuck in reprojection because from here on, DCS will be CPU limited. 

So, the benefit is really cranking up all settings to high (that said, with a varjo aero!). Is it worth 2000 bucks? For some of us yes, for some not. 

I didn't expect that kind of impact myself and am torn right now whether to buy or not. With a mild tendency to say yes. Because from now on and until the CPU limit isn't addressed, this card will serve you a long time unless new headsets are coming out with a lot higher resolution. I really can't see this coming in the next years. If you look at the current new releases (Pico, varjo, etc) they all have the more or less same resolution and are opting in for better features like wireless, eye tracking and so on. As technology evolves (eye tracking, Vulcan) the demand will rather decrease than increase. 

But that's just my two cents here. 

  • Like 4

GeForce RTX 4090 Founders Edition - AMD Ryzen 7 5800X3D - 64Gb RAM - Win11 - HP Reverb G1 - Thrustmaster Warthog HOTAS (40cm extension) - VKB Sim T-Rudder MKIV Pedals

Link to comment
Share on other sites

I'm running an i9 10900k and went from a 3090 to a 4090, on a 4k 120hz panel. In DCS, there was no noticeable improvement so far.  I tried a bunch of games, and went into the settings to try to max them out, but I noticed they were already maxed out.

Some games did show improvements - RDR2 is much better, MSFS is hard to tell but seems better, the battlefield games already ran great on the 3090.

One improvement I noticed for sure is noise.  I was running an EVGA XC3 3090, and it used to get really loud.  My MSI 4090 fans are set to auto, and they haven't gone over 30% yet.  It pumps out a lot of heat, but the cooler is really efficient.  I'd really hope so, given the insane size.  I had to literally cut holes in my case and mount internal case fans on the outside to fit the thing in.  It's gigantic.

run come save me

Link to comment
Share on other sites

10 minutes ago, Scofflaw said:

I'm running an i9 10900k and went from a 3090 to a 4090, on a 4k 120hz panel. In DCS, there was no noticeable improvement so far.  I tried a bunch of games, and went into the settings to try to max them out, but I noticed they were already maxed out.

Some games did show improvements - RDR2 is much better, MSFS is hard to tell but seems better, the battlefield games already ran great on the 3090.

One improvement I noticed for sure is noise.  I was running an EVGA XC3 3090, and it used to get really loud.  My MSI 4090 fans are set to auto, and they haven't gone over 30% yet.  It pumps out a lot of heat, but the cooler is really efficient.  I'd really hope so, given the insane size.  I had to literally cut holes in my case and mount internal case fans on the outside to fit the thing in.  It's gigantic.

I am no computer except but it seems like you are CPU bottlenecked.

Link to comment
Share on other sites

32 minutes ago, Rubber Biscuit said:

I am no computer except but it seems like you are CPU bottlenecked.

Possible, but I doubt it.  My water-cooled 10900k is overclocked to 5.2GHz on all 10 cores.  Aside from benchmarks, I've never had any game or application push my CPU anywhere near 100%.  I can do a stable 5.3ghz, but it'll start to hit gnarly temps when I do that.

The only major improvement I could do on this setup is go from 32 to 64GB of RAM, but I was saving that for my next build.

run come save me

Link to comment
Share on other sites

If your GPU does not show close to 100% utilisation when gaming on a 2D monitor, and you are not at the vsync limit, then yes, you are most likely limited by CPU. 

With CPU it's different, basically no game will use it to 100% and looking at cpu utilisation is mostly pointless.

In MSFS switching from rtx3080 I saw 40% increase on ultra settings in one spot i tested, and 100% increase in another scenario that is not cpu limited. Except I have 5800x3d, while 10900k performance in this game is comparable with 5600x/5900x

de9d32279b43b2fa8871575e3baf15182cf35ba3

b91b0ee1b53e621d8b0bd39604ebf8774b4e12a3

  • Like 1
  • Thanks 1

Hardware: VPForce Rhino, FSSB R3 Ultra, Virpil T-50CM, Hotas Warthog, Winwing F15EX, Slaw Rudder, GVL224 Trio Throttle, Thrustmaster MFDs, Saitek Trim wheel, Trackir 5, Quest Pro

Link to comment
Share on other sites

5 hours ago, Scofflaw said:

Possible, but I doubt it.  My water-cooled 10900k is overclocked to 5.2GHz on all 10 cores.  Aside from benchmarks, I've never had any game or application push my CPU anywhere near 100%.  I can do a stable 5.3ghz, but it'll start to hit gnarly temps when I do that.

Checking CPU frame times is the real test I believe. Most games use few cores and their use get shared out amongst all cores so the overall percentage CPU usage always appears low, even if it is the performance bottleneck.

  • Like 2
  • Thanks 1

AMD 5800X3D · MSI 4080 · Asus ROG Strix B550 Gaming  · HP Reverb Pro · 1Tb M.2 NVMe, 32Gb Corsair Vengence 3600MHz DDR4 · Windows 11 · Thrustmaster TPR Pedals · VIRPIL T-50CM3 Base, Alpha Prime R. VIRPIL VPC Rotor TCS Base. JetSeat

Link to comment
Share on other sites

1 hour ago, AirMeister said:

Well : 
***
SO DO GAMES USE MULTIPLE CORES? 
The short answer is yes. Almost all of the games these days, even the simplest indie games, make use of multiple cores.
***

Only the very low overhead 'esports' type of games use few (2 to 4) cores. These are games where you typically have 300+ fps. 
As for any other game, 6 to 8 cores is typical. 


https://pcguide101.com/cpu/do-games-use-multiple-cores/

In the scheme of the poster's 10900k 10 cores, the games in the article you quoted all use less than all of them (significantly in most cases) so the CPU will never show 100% utilisation. That was the point, whether the wording is few, half or not all cores used.

  • Like 1
  • Thanks 1

AMD 5800X3D · MSI 4080 · Asus ROG Strix B550 Gaming  · HP Reverb Pro · 1Tb M.2 NVMe, 32Gb Corsair Vengence 3600MHz DDR4 · Windows 11 · Thrustmaster TPR Pedals · VIRPIL T-50CM3 Base, Alpha Prime R. VIRPIL VPC Rotor TCS Base. JetSeat

Link to comment
Share on other sites

On 10/16/2022 at 8:25 PM, Scofflaw said:

Possible, but I doubt it.  My water-cooled 10900k is overclocked to 5.2GHz on all 10 cores.  Aside from benchmarks, I've never had any game or application push my CPU anywhere near 100%.  I can do a stable 5.3ghz, but it'll start to hit gnarly temps when I do that.

The only major improvement I could do on this setup is go from 32 to 64GB of RAM, but I was saving that for my next build.

You won’t because most games don’t use over 6 cores. What you’re seeing is total CPU usage taking all cores into account.

  • Thanks 1
Link to comment
Share on other sites

Install CapframeX, run your "standard" benchmark flight whist capturing. When you check the results look at the CPU thread utilisation and GPU util and you'll be able to work out easily which component is the bottleneck. My 12900K is my bottleneck in both DCS and the "other" SIM since my GPU upgrade.

  • Like 2
  • Thanks 1

12900KF | Maximus Hero Z690 | ASUS 4090 TUF OC | 64GB DDR5 5200 | DCS on 2TB NVMe | WarBRD+Warthog Stick | CM3 | TM TPR's | Varjo Aero

Link to comment
Share on other sites

17 hours ago, zildac said:

Install CapframeX, run your "standard" benchmark flight whist capturing. When you check the results look at the CPU thread utilisation and GPU util and you'll be able to work out easily which component is the bottleneck. My 12900K is my bottleneck in both DCS and the "other" SIM since my GPU upgrade.

This is good info thanks for posting.

this will help us all when building up around a new RTX-4090

what percentage of load did your gpu get to.

 

Link to comment
Share on other sites

Thanks for the video!

This is about what I noticed going from a 2080Ti to a 3090. You dont get significantly better FPS but you get lots of overhead so you can turn up your visual settings and pixel density. In the end you get way better clarity and visuals at about the same FPS you had before. Usually the overall performance is more consistent less stutter. My goal was to set graphics setting that gave me about 80 to 90 GPU utilization. Not surprised that nothing has changed with the 40 series.  Better graphics at about the same FPS. I wonder how much vulkan and multi core will change things if that ever happens?

  • Like 2
  • Thanks 1
Link to comment
Share on other sites

On 10/13/2022 at 12:45 AM, Kiwispirits said:

Any chance you could post a shot of the screen settings page from DCS that you used when you got 90 fps please?

 

just played hoggit multiplayer server. started in hangar.. never dipped below 60 fps

i had 70-90 most of the time..

i am 5800x3d and 4090

dcs set.jpg

  • Thanks 1

find me on steam! username: Hannibal_A101A

http://steamcommunity.com/profiles/76561197969447179

Link to comment
Share on other sites

5 hours ago, hannibal said:

just played hoggit multiplayer server. started in hangar.. never dipped below 60 fps

i had 70-90 most of the time..

i am 5800x3d and 4090

dcs set.jpg

In Italy we say "done 30, let's do 31" 😀
Can You pls also posting the VR tab and eventually Nvidia Control Panel settings?
have You setting up something in the openxr (FSR/NIS/Foveted...)?
So we could have the total picture.
Thx.

 

Link to comment
Share on other sites

En 21/10/2022 a las 5:47, Rubber Biscuit dijo:

Looks like the CPU is having trouble keeping up.

a little optimization could provide a huge boost to fps

Thanks! Sure. Using the default presets I try to make things easier so that everyone can do the benchmark at home easily.

Once Varjo fixes the stuttering issue with the 4090 I'll start fine-tuning the graphics settings to get the best possible experience but until then I'll keep benchmarking and testing. 🙂

Here's another one (Apache + Marianas) comparing 2 different resolutions of the VARJO AERO. 35 and 39 PPD (Pixels Per Degree) corresponding to 3140x2692 and 4148x3556 respectively at 100% of SteamVR.

 

  • Thanks 3
Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...