5090 game performance on par with 4090.

[Rogueywon]

Ive actually already DDU’d and its made no difference. Strangely Ive also got a small increase from the new drivers but that is only in 3DMark benchmarks. I get 14822 in Steel Nomad now, up from about 14300. It just seems to be games that I get a lack of performance.

I forgot to mention in the OP that I only got 150-200 fps in Mech Warriors 5: Mercenaries, thats with RT off, a 4090 on youtube paired with a 5800x3d gets 300-400. Literally double my fps.

I wouldn't read too much into the MW5 Mercs point. It's been bugged to hell for a while, I think via drivers, and often refuses to make full use of GPU or CPU (and it's not choking on CPU threads either).

 

[Daytrader]

I reckon you just have graphics settings set to high in nvidia control panel or app, that will give you lower game benchmarks.

 

[Rroff]

maybe the degradation problem that Intel had over the last two years

Doesn't usually reduce performance, most commonly manifests as a crash when the game is compiling shaders or decompressing files, if it is really far gone you'll randomly get memory access related BSODs when trying to boot into Windows.

 

[no_1_dave]

So much incorrect information in here. On one hand I'm surprised because it's OCUK which is normally pretty solid, but on the other hand I guess it's still the internet...

At 4K with a 4090/5090 there is almost no difference in gaming performance between the latest X3D CPUs and Intel CPUs, even going back several generations.
In some games there is a small difference, mainly to minimum fps, but in the majority there is no difference at all.

Hardware Unboxed even made an April Fools video on it, teasing people who say they review CPUs incorrectly because they're always benched at 1080p.
Even their joke video proves there is no tangible difference, yet on here people are claiming this will cost the OP 20%-40%..?

OP - There is nothing wrong with your CPU. I don't know what the problem could be, but it's nothing to do with your 14700k vs modern X3D CPU

Edit - Check and update your motherboard bios and check which cores your games are running on. I had an old Asus Prime board which kept running games on my e-cores instead of p-cores which caused similar problems to what you describe. I fixed it by updating the bios and running windows update which fixed core allocation in games. The prime board was still useless though and kept trying to push my ffmpeg encoding to e-cores so I had to disable them until changing to a Gigabyte board and it's been fine ever since

 

[Clarence Boddicker]

OP - There is nothing wrong with your CPU. I don't know what the problem could be, but it's nothing to do with your 14700k vs modern X3D CPU

Edit - Check and update your motherboard bios and check which cores your games are running on. I had an old Asus Prime board which kept running games on my e-cores instead of p-cores which caused similar problems to what you describe. I fixed it by updating the bios and running windows update which fixed core allocation in games. The prime board was still useless though and kept trying to push my ffmpeg encoding to e-cores so I had to disable them until changing to a Gigabyte board and it's been fine ever since

Ok thanks, just had Quick Look at resource manager and HWmonitor while game was running and p cores were between about 60 and 80% while e were fluctuating between 40 and 60%, How did you identify which particular cores your games were running on?

 

[Joxeon]

Is the GPU pegged at 99-100% while gaming?

 

[no_1_dave]

Ok thanks, just had Quick Look at resource manager and HWmonitor while game was running and p cores were between about 60 and 80% while e were fluctuating between 40 and 60%, How did you identify which particular cores your games were running on?

My e-cores had more usage than my p-cores when I was checking CPU usage.
Try disabling e-cores in the bios so it only has the option for p-cores, then test and see how games run

 

[Clarence Boddicker]

Is the GPU pegged at 99-100% while gaming?

No, it fluctuates between 70 and 100%.

My e-cores had more usage than my p-cores when I was checking CPU usage.
Try disabling e-cores in the bios so it only has the option for p-cores, then test and see how games run

Ok, my p-cores seem to have about 20-30% more usage than the e-cores so not sure if that is the problem then. I've got some faster RAM on the way so I’ll see what sort of difference that makes and then give disabling the e-cores a go.

 

[cp_997]

I’ve got a 13700K which I’ve had for over 2 years now and had a 4090 previous before getting my 5090 FE in Feb. I’ve found my benchmarks are within a frame or 2 of even the 9800X3D (@ 4K) and are all round faster than my 4090 was. I have a Z790 asus motherboard on latest bios and run with 5600 ram. I’ve checked GPU-Z and it’s running PCIe 5. Don’t think anyone has asked but what is your PSU? Are you running air cooler on CPU or AIO?

 

[Clarence Boddicker]

I’ve got a 13700K which I’ve had for over 2 years now and had a 4090 previous before getting my 5090 FE in Feb. I’ve found my benchmarks are within a frame or 2 of even the 9800X3D (@ 4K) and are all round faster than my 4090 was. I have a Z790 asus motherboard on latest bios and run with 5600 ram. I’ve checked GPU-Z and it’s running PCIe 5. Don’t think anyone has asked but what is your PSU? Are you running air cooler on CPU or AIO?

My psu is an MSI MPG 1250 and Ive got a 360 AIO, PCIe 4. The strange thing is that my 3DMark benchmarks are all inline with other 5090’s. It’s just games where I’m way behind in performance. Looking at more videos on YouTube, I think I’m actually behind 4090’s in performance in game, but ahead in 3DMark benchmarks.

 

[cp_997]

My psu is an MSI MPG 1250 and Ive got a 360 AIO, PCIe 4. The strange thing is that my 3DMark benchmarks are all inline with other 5090’s. It’s just games where I’m way behind in performance. Looking at more videos on YouTube, I think I’m actually behind 4090’s in performance in game, but ahead in 3DMark benchmarks.

PSU is higher rated than mine as I’ve got 1000W, and PCIe 4/5 makes little difference in reality. I don’t suppose you’ve got access to someone else’s 5090, or they could try yours in their machine given they’re still a bit like hens teeth (expensive hens teeth at that!). Only variable here seems to be the 5090 itself if everything was better with the 4090!

 

[Rroff]

I assume you've already checked but the GPU is actually running full PCI-e speed in GPU-z? (might have to do the render test to get it to fully power up the link).

 

[no_1_dave]

I've got some faster RAM on the way so I’ll see what sort of difference that makes and then give disabling the e-cores a go.

It's just a bios toggle, very quick and easy to test while waiting for the ram

 

[Clarence Boddicker]

PSU is higher rated than mine as I’ve got 1000W, and PCIe 4/5 makes little difference in reality. I don’t suppose you’ve got access to someone else’s 5090, or they could try yours in their machine given they’re still a bit like hens teeth (expensive hens teeth at that!). Only variable here seems to be the 5090 itself if everything was better with the 4090!

Yeah its baffling, I’d think there was something with the GPU if it wasn’t for the decent benchmark scores.

I assume you've already checked but the GPU is actually running full PCI-e speed in GPU-z? (might have to do the render test to get it to fully power up the link).

When I first installed it, it wasn’t, I was getting very poor benchmark scores and constant crashes. When I checked the PCI-e Gen in the bios it was set to Auto so I changed it to Gen 4 and it’s been stable ever since with no crashes, benchmarks jumped up to where they should be as well.

It's just a bios toggle, very quick and easy to test while waiting for the ram

Ok I’ll give it a try tomorrow.

 

[Grim5]

I assume you've already checked but the GPU is actually running full PCI-e speed in GPU-z? (might have to do the render test to get it to fully power up the link).

Yeah PCIE speed ain't gonna make up 20% perf loss, 1-3% maybe but not 20%

 

[scubes]

could his card have missing rops?

 

[Varys]

Have you run Timespy - It gives separate GPU and CPU scores so you can then compare with similar systems

 

[Joxeon]

No, it fluctuates between 70 and 100%.

Does it do that when stationary or just when moving around in game?

 

[docsonic]

Have you tried a fresh windows install yet as suggested earlier in the thread? If you have a spare drive knocking around its a fast way to discount anomalies within your current install.

 

[Clarence Boddicker]

could his card have missing rops?

No, none missing.

Have you run Timespy - It gives separate GPU and CPU scores so you can then compare with similar systems

I have, no issue with the cpu score.

Does it do that when stationary or just when moving around in game?

I think that its more stable when stationary, Ill have to check that later though to be sure.

Have you tried a fresh windows install yet as suggested earlier in the thread? If you have a spare drive knocking around it’s a fast way to discount anomalies within your current install.

No, I havent had the time to yet and its not something Ive done before so I’ll have to do a bit of research on how to go about it.