5080RTX FE Performance normal?

[SlowBlinkCat]

I am not sure if I should be getting the FPS that I should be with this card. I built a new pc with the following specs:

CPU: 9950X3D
GPU: Nvidia 5080RTX FE
RAM: Flare X5 Series 64GB (2 x 32GB) DDR5-6000 PC5-48000 CL30
MOBO:MSI X870E MAG TOMAHAWK WIFI AMD AM5
SSD: 990 PRO 4TB Samsung
PSU: RMx SHIFT Series RM1200x 1200 Watt 80 PLUS

I mostly play COD and on Warzone Verdansk I am getting 210-280fps around the map (Playing on 1440p LOW settings, frame-gen OFF, no DLSS). I have a friend who has a 7800x3d and 4070 Super, and the same settings as me (1440p) who is getting the same or more fps than I am. I'm not sure if its because of their CPU and if mine is more balanced towards non-gaming tasks. I don't know much about CCDs or core parking.

I have EXPO enabled, the latest nvidia drivers (581.15), and the latest AMD Chipset software (7.06.02.123) installed. I also checked on GPU-Z and it shows my bus interface for my GPU running at PCIe x 16 5.0 @ x16 5.0 (when I am playing a game), when idle its @ x16 1.1.

I do not have PBO enabled or any overclock for my CPU.

Does this sound correct performance-wise? Is there another game that I should play to correctly test or should I do a benchmark test with 3dmark?

Edit* Below is 3DMark demo benchmark using Time Spy

 

[GoldenTiger]

should I do a benchmark test with 3dmark?

Yes.

 

[KickAssCop]

Seems slow.

You should be getting something like 300-315 fps in the built in benchmark using max settings with either DLSS or FSR and no FG.

 

[GoldenTiger]

Seems slow.

You should be getting something like 300-315 fps in the built in benchmark using max settings with either DLSS or FSR and no FG.

He's just eyeballing it during actual gameplay. Yes, posting the benchmark and a 3dmark would be helpful.

 

[SlowBlinkCat]

Here is a time spy benchmark from 3dmark demo, I can't do the in-game COD benchmark since I don't have Black Ops 6 installed/purchased, and I don't see a benchmark tool within Warzone.

 

[SlowBlinkCat]

Seems slow.

You should be getting something like 300-315 fps in the built in benchmark using max settings with either DLSS or FSR and no FG.

Are those frames from the benchmark for multiplayer? My range I posted is for Warzone. What do you typically get on Warzone?

 

[KickAssCop]

Are those frames from the benchmark for multiplayer? My range I posted is for Warzone. What do you typically get on Warzone?

Benchmark.
Those are not my frames but a friend of mine.
I run maxed out with a 5090 on 4K 240.

 

[SlowBlinkCat]

Benchmark.
Those are not my frames but a friend of mine.
I run maxed out with a 5090 on 4K 240.

Got it, yea I would think I get those same frames if I were to benchmark multiplayer since usually multiplayer frames are higher than warzone

 

[jobert]

Strange your cpu score is a little less than my 12700k gets in Time Spy.

 

[SlowBlinkCat]

My CCD1 cores do not park when I have the game running. I have CPPC Dynamic Preferred Core set to "Driver" in the BIOS, and I also have re-installed the AMD chipset drivers. I also have AMD 3D V-Cache Performance Optimizer service running in task manager. I had Power plan on balanced, Game mode and Game bar both ON.

but after I close the game, CCD0 goes to sleep as it should I assume:

To fix the issue, I used Process Lasso to set Call of Duty to CCD0 (0-15) and that boosted my fps. I now get 250-320fps. For this, I disabled Game bar and game mode on Windows 11. And I set CPPC Dynamic Preferred Core to "Frequency" in the BIOS.

Why doesn't it work without Process lasso though?

 

[KickAssCop]

I don't use project lasso or anything else. All runs fine. Very strange that you are having to rely on it.

 

[SlowBlinkCat]

I don't use project lasso or anything else. All runs fine. Very strange that you are having to rely on it.

you have a 5090 though

 

[KickAssCop]

Graphics card should not be a factor on how your cores get parked. That is managed by windows and drivers. So there is something fishy with your installs causing it to not work as advertised.

My suggestion would be to ensure that you have latest MSI bios installed. Additionally, you ensure that chipset drivers are installed correctly (uninstall and reinstall). Final thing I would do is confirm the Xbox game bar etc. is working as it should. I know you already said windows is in balanced mode so probably don't need anything to do there.

I would also run DISM command to restore image online in case of any windows issues you might not be aware of.

That is the extent of my knowledge about this. I myself moved from a 9800X3D to a 9950X3D and my windows has seen processors 5600X, 5900X, 5800X3D, 7800X3D, 9800X3D and now 9950X3D without a need for reinstall. Notice I didn't do a windows reinstall when I swapped out motherboards from X570E to X670E also. I have also swapped out ram and HDD components multiple times. FYI.

 

[SlowBlinkCat]

Graphics card should not be a factor on how your cores get parked. That is managed by windows and drivers. So there is something fishy with your installs causing it to not work as advertised.

My suggestion would be to ensure that you have latest MSI bios installed. Additionally, you ensure that chipset drivers are installed correctly (uninstall and reinstall). Final thing I would do is confirm the Xbox game bar etc. is working as it should. I know you already said windows is in balanced mode so probably don't need anything to do there.

I would also run DISM command to restore image online in case of any windows issues you might not be aware of.

That is the extent of my knowledge about this. I myself moved from a 9800X3D to a 9950X3D and my windows has seen processors 5600X, 5900X, 5800X3D, 7800X3D, 9800X3D and now 9950X3D without a need for reinstall. Notice I didn't do a windows reinstall when I swapped out motherboards from X570E to X670E also. I have also swapped out ram and HDD components multiple times. FYI.

I meant you are probably getting higher frames regardless. So you have checked resource monitor or AMD Ryzen Master and your CCD1 goes to sleep when gaming on your 9950x3d?

The only reason I questioned my performance was after I compared to my friend's rig otherwise I would have never really known.

Even after Process lasso, CCD1 still does not go to sleep entirely since there are processes running in the background utilizing both CCDs, but I still get that FPS boost I am looking for. In other words how can you know you won't get better performance until you try Process Lasso?

 

[KickAssCop]

I meant you are probably getting higher frames regardless. So you have checked resource monitor or AMD Ryzen Master and your CCD1 goes to sleep when gaming on your 9950x3d?

The only reason I questioned my performance was after I compared to my friend's rig otherwise I would have never really known.

Even after Process lasso, CCD1 still does not go to sleep entirely since there are processes running in the background utilizing both CCDs, but I still get that FPS boost I am looking for. In other words how can you know you won't get better performance until you try Process Lasso?

That's a fair callout. I can download it and try. If I have time I will let you know how that worked out.

 

[Absalom]

Other than the CPU, OP & I have very similar specs. I tried to download Timespy from Steam but it threw up a "No licenses" error. Guess they ran out of licenses.

If there's another benchmark that's not 10s or 100s of GBs to download, I'll run it and we can compare numbers.

 

[KickAssCop]

Technically he has average in his screenshot and he is below that average.

 

[Absalom]

Technically he has average in his screenshot and he is below that average.

1% difference. Some might call that "margin of error".

 

[KickAssCop]

1% difference. Some might call that "margin of error".

Yes can be simply because his card silicon didn't keep the clocks high enough during that run.

 

[T4rd]

This is why I would still take a 9800X3D over a 9900/9950X3D unless you have a non-gaming/productivity need for the extra cores. It's not worth the CCD/parking hassle if you're getting noticeably worse performance while also running hotter and clocking lower than the single CCD chip. It may be less of an issue these days with the current scheduler (not sure, I haven't researched it recently), but even if it only happens 5% of the time, that would be too much for me.

 

[KickAssCop]

This is why I would still take a 9800X3D over a 9900/9950X3D unless you have a non-gaming/productivity need for the extra cores. It's not worth the CCD/parking hassle if you're getting noticeably worse performance while also running hotter and clocking lower than the single CCD chip. It may be less of an issue these days with the current scheduler (not sure, I haven't researched it recently), but even if it only happens 5% of the time, that would be too much for me.

I was on the same boat. But honestly, I have felt no difference in gaming which is my primary use case. I love that shader compile doesn’t take ages and the flexibility I have for doing productivity tasks big or small. So for me it has been a totally worth it upgrade.

 

[T4rd]

I was on the same boat. But honestly, I have felt no difference in gaming which is my primary use case. I love that shader compile doesn’t take ages and the flexibility I have for doing productivity tasks big or small. So for me it has been a totally worth it upgrade.

Yeah, but you could probably move to a 9600X and realistically not notice much of a difference in-game because you're still mostly GPU limited in games and getting 120+ fps. But if I had a 240hz+ monitor and I checked my frames to see I'm not getting close to what I should be on a 9800X3D while on a 9950X3D or sporadic frame pacing in a game not known for it, I would be pretty annoyed that I paid more for worse performance or had to spend my limited gaming time troubleshooting and trying to park cores to one CCD. I paid for the whole monitor and want to use the whole monitor, hah.
So I'll take waiting an extra min or two of a one-time shader compile for the comfort and convenience that I'll be getting optimal performance no matter what, or at least one less factor to troubleshoot if I happen to notice anything.

 

[KickAssCop]

Yes of course lots of choices. Get what suits you best. For me 9950X3D has been great.

 

[Absalom]

I *just* got this 5080 FE to upgrade a 3080 and I can't tell a difference. Seems overkill for 3440x1440@100Hz, but getting a unicorn like the 5080 FE I couldn't pass up. I need to stop playing games that are 6 years old. Got back into The Division 2, but man that game still looks great!

 

[SlowBlinkCat]

That's a fair callout. I can download it and try. If I have time I will let you know how that worked out.

I need to ask you though, do you have the AMD Adrenalin software installed? I read that some people updated their chipset through the software, but you should only download the software if you have a Radeon graphics card, am I right?

I have some good news. I removed Process Lasso to try one more time without it. BIOS CPPC set to "Driver", Game Mode ON, Game bar ON, Power plan on Balanced, AMD Chipset software to latest version.

I ended up opening Game Bar and enabling "Remember this is a Game" while having Warzone running in the background. That instantly boosted my frames by like 50. Ridiculous.

It's still not as much as Process Lasso gave me though and I do notice the cores are not entirely going to sleep on AMD Ryzen Master. Strange.

EDIT** CCD1 is finally parking. Looks like I had to switch the game to windowed mode and make sure I did not click out of the window to see the true result. All is well now. Process Lasso will give a small boost but I'm perfectly fine running without it.

 

[KickAssCop]

Cool glad you got it sorted. I delete amd radeon drivers from my pc with prejudice. I even have onboard disabled in bios and windows because somehow it keeps showing up as GPU0 in MSI AB.

I didn’t get to try process lasso but I have benches from both 9800x3D and 9950X3D on same pc and in COD my 9950X3D was about 2-3 fps faster possibly due to more aggressive ram tuning.

 

[SlowBlinkCat]

Cool glad you got it sorted. I delete amd radeon drivers from my pc with prejudice. I even have onboard disabled in bios and windows because somehow it keeps showing up as GPU0 in MSI AB.

I didn’t get to try process lasso but I have benches from both 9800x3D and 9950X3D on same pc and in COD my 9950X3D was about 2-3 fps faster possibly due to more aggressive ram tuning.

I noticed when I did 3dmark benchmark it also shows AMD Radeon Graphics below the 5080, I don't have the drivers installed though. It also shows it in HWMonitor. This shouldn't impact performance though should it?

 

[KickAssCop]

I noticed when I did 3dmark benchmark it also shows AMD Radeon Graphics below the 5080, I don't have the drivers installed though. It also shows it in HWMonitor. This shouldn't impact performance though should it?

No it does not but for good measure disable it in bios and especially windows.

 

[toddw]

Kinda got lost in this thread but to answer the orig OP question, that timespy score does not look abnormal for a stock FE 5080.

 

[SlowBlinkCat]

Kinda got lost in this thread but to answer the orig OP question, that timespy score does not look abnormal for a stock FE 5080.

Yes, running another benchmark yields just slightly above average so performance wise I am good. It was Xbox Game Bar all along ("Remember this is a Game" option which allows the CPU to park the regular CCD).

 

[SlowBlinkCat]

Seems slow.

You should be getting something like 300-315 fps in the built in benchmark using max settings with either DLSS or FSR and no FG.

So I see your friend's benchmark screenshot is from early August, I just ran one now and I don't get near those frames. This is with frame gen off and using Fidelity CAS. I noticed my VRAM usage is much higher compared to your screenshot.

My GPU drivers are newer, but that still doesn't explain the CPU performance. The 9950x3d should be about the same as the 9800x3d.

Is your friend overclocking anything?

 

[KickAssCop]

I will have to check. Did you try with process lasso also?

 

[SlowBlinkCat]

I will have to check. Did you try with process lasso also?

I did not. It would be a little higher though for sure. I uninstalled Process Lasso completely, just didn't want to rely on a 3rd party software running everytime.

I also found this gameplay video on youtube of someone with a 9800x3d (with PBO on) and 5080rtx on Warzone, and I get around the same FPS as him with the same settings, mines is maybe 10fps less.

View: https://www.youtube.com/watch?v=TdWhdg8C5z8&ab_channel=Hakaluk

So compared to this I'm getting just about the same performance as a 9800x3d. I'm just curious as to how your friend has 310fps benchmark on multiplayer.

 

[T4rd]

I did not. It would be a little higher though for sure. I uninstalled Process Lasso completely, just didn't want to rely on a 3rd party software running everytime.

I also found this gameplay video on youtube of someone with a 9800x3d (with PBO on) and 5080rtx on Warzone, and I get around the same FPS as him with the same settings, mines is maybe 10fps less.

View: https://www.youtube.com/watch?v=TdWhdg8C5z8&ab_channel=Hakaluk

So compared to this I'm getting just about the same performance as a 9800x3d. I'm just curious as to how your friend has 310fps benchmark on multiplayer.

Yeah, I would expect the 9800X3D to run marginally faster due to it being able to clock a bit higher.

 

[KickAssCop]

I did not. It would be a little higher though for sure. I uninstalled Process Lasso completely, just didn't want to rely on a 3rd party software running everytime.

I also found this gameplay video on youtube of someone with a 9800x3d (with PBO on) and 5080rtx on Warzone, and I get around the same FPS as him with the same settings, mines is maybe 10fps less.

View: https://www.youtube.com/watch?v=TdWhdg8C5z8&ab_channel=Hakaluk

So compared to this I'm getting just about the same performance as a 9800x3d. I'm just curious as to how your friend has 310fps benchmark on multiplayer.

Can’t connect with him rn. Maybe he turned down settings or something. But he was definitely overclocking.

 

[SlowBlinkCat]

Also for anyone finding this in the future, setting the CPU Renderer Count in the call of duty config file to "8" instead of "16" for the 9950x3d boosts frames by another 30-40fps by further avoiding any performance mishaps with the non-3D V-Cache cores. I'm now getting 230-330fps around warzone with a stock 9950x3d/5080rtx combo

 

[KickAssCop]

Also for anyone finding this in the future, setting the CPU Renderer Count in the call of duty config file to "8" instead of "16" for the 9950x3d boosts frames by another 30-40fps by further avoiding any performance mishaps with the non-3D V-Cache cores. I'm now getting 230-330fps around warzone with a stock 9950x3d/5080rtx combo

This is interesting. Wondering if I should attempt this but also don't want to mess anything up (since COD is finicky on game pass version). Plus don't want to get banned or some crap. I already hit 240 fps at 4K.

 

[SlowBlinkCat]

This is interesting. Wondering if I should attempt this but also don't want to mess anything up (since COD is finicky on game pass version). Plus don't want to get banned or some crap. I already hit 240 fps at 4K.

Wouldn't it be the proper thing to do though? A 9800x3d would have it set to 8 or 7, and so a 9950x3d should have it too to reach similar performance.

 

[SlowBlinkCat]

This is interesting. Wondering if I should attempt this but also don't want to mess anything up (since COD is finicky on game pass version). Plus don't want to get banned or some crap. I already hit 240 fps at 4K.

I was able to snag a 5090 at MSRP! Updated benchmark with everything at stock.

 

[KickAssCop]

I was able to snag a 5090 at MSRP! Updated benchmark with everything at stock.

Why you run fidelity CAS BS on a 5090 bruh?