RTX 3070 at 10600K cant get above 100-120 FPS in Warzone and Underperforms in some games

[BillNaaai]

Not really the intention sorry if it came across that way. I’ve read everyone response and every fix suggested I have done so I’m not sure what you mean when you say I’m not reading? Very much Open to any advice that may help

Just comes across that you’re being defensive saying constantly your gpu should be better at every step of the way

 

[CLAM]

In Warzone on all settings the 10600k doesn't even beat the 9700k nvmd the 9900k.
Sure the performance is close to that of a 9900k, but by no means does it 'beat it' as you stated.

However, if you don't manage to sort your issue out of low fps, feel free to send the card my way (If you are in Cpt) and I can run some tests on it and compare it to my 3070 to see if you got a lemon.

I’m this video see the frame rate with the 10600k and a 2080TI at 1440p it’s getting like 160. I guess maybe those frames I’m looking for will come with faster ram and a Z board?

 

[CLAM]

Just comes across that you’re being defensive saying constantly your gpu should be better at every step of the way

Apologies man definitely not the intention. It is an RTX 3070 and there are many benchmarks out there which support what I’m saying. So not trying to discredit people suggestions but rather stating why I don’t think something is the case.

 

[CLAM]

Apologies I stand corrected.

I'm in Joburg so might be difficult to take you up on the offer but thanks

 

[∀Z‾soʇɐuɐɥ┴]

I'm in Joburg so might be difficult to take you up on the offer but thanks

Im sure there are a few nice members that side that would be willing to check if your gpu is a lemon as well.
But yes, I think the other guys with their thinking around XMP not being enabled might be onto something.

As a test I will put my 3070 in my younger bro's pc later when he wakes up (he is running a 9700k on a z490 with xmp enabled on ram) and see how it performs in Warzone to get a good idea of where about it should be performing seeing that in the vid I linked the 9700k and the 10600k performed quite close to each other, albeit with a 2080ti.

 

[CLAM]

If you would be willing to do that I would be hugely grateful!
Maybe clock his ram down to 2666 to get it a bit closer to where my system is.

 

[CLAM]

Im sure there are a few nice members that side that would be willing to check if your gpu is a lemon as well.
But yes, I think the other guys with their thinking around XMP not being enabled might be onto something.

As a test I will put my 3070 in my younger bro's pc later when he wakes up (he is running a 9700k on a z490 with xmp enabled on ram) and see how it performs in Warzone to get a good idea of where about it should be performing seeing that in the vid I linked the 9700k and the 10600k performed quite close to each other, albeit with a 2080ti.

Do you play warzone at all, how is your card performing in your system? (although I know it destroys mine lol).

 

[∀Z‾soʇɐuɐɥ┴]

Do you play warzone at all, how is your card performing in your system? (although I know it destroys mine lol).

I did play it at first, but prefer the normal multiplayer to warzone.
As such I haven't even installed Battle.net on my pc or updated the game in about 6 or 7 months
Last time I played it I still had a 5700xt

My younger bro plays it almost daily however with my cousin.

 

[Petester]

I've done that and GPU is definitely underutilizing.

If the GPU is underutilized then thats one of the problems. Is this something you have observed in data via MSI afterburner or are you just presuming it is being underutilised ? . It's something you can monitor, so it shouldn't be guesswork. Once you have determined that via the data, then you are halfway to finding a solution.

 

[∀Z‾soʇɐuɐɥ┴]

If the GPU is underutilized then thats one of the problems. Is this something you have observed in data via MSI afterburner or are you just presuming it is being underutilised ? . It's something you can monitor, so it shouldn't be guesswork. Once you have determined that via the data, then you are halfway to finding a solution.

Probably used task manager, though the utilization does drop if you alt+tab to task manager and a game gets pushed to the background

 

[Petester]

Probably used task manager, though the utilization does drop if you alt+tab to task manager and a game gets pushed to the background

If he has the OSD on while gaming he should be able to see all the relevant data

 

[∀Z‾soʇɐuɐɥ┴]

I did play it at first, but prefer the normal multiplayer to warzone.
As such I haven't even installed Battle.net on my pc or updated the game in about 6 or 7 months
Last time I played it I still had a 5700xt

My younger bro plays it almost daily however with my cousin.

Out of interest sake decided to update Warzone on my pc now, will see how it performs in 20-30mins after it has finished it's 90gb update.

 

[Petester]

Also max out your power limit in Afterburner for what it's worth.

 

[Mew_Tw0]

Go into your Nvidia experience settings. Enable the overlay, then go into the settings for the overlay to enable experimental features. Then in the overlay, enable the performance stats.

While in game, press ctrl+r and a lil window will pop up that shows all the info you will need. Usage and Temps for cpu and gpu, framerate, average lows, etc.

Also you might take into account that different games perform differently in terms of what cpu you playing on and what gpu.

The rtx 3070 has been a great card for me with my 1800x. I saw double frame rate increases across most games when comparing it to my 1070ti.

The best way to truly test, is a fresh install of windows and the game, nothing else. And test. Also you might try older versions of the Nvidia drivers.

 

[CLAM]

If the GPU is underutilized then thats one of the problems. Is this something you have observed in data via MSI afterburner or are you just presuming it is being underutilised ? . It's something you can monitor, so it shouldn't be guesswork. Once you have determined that via the data, then you are halfway to finding a solution.

I have a sidebar GPU and CPU monitor as well as afterburner running on my 2nd monitor so I definitely know it's underutilizing.

 

[CLAM]

Go into your Nvidia experience settings. Enable the overlay, then go into the settings for the overlay to enable experimental features. Then in the overlay, enable the performance stats.

While in game, press ctrl+r and a lil window will pop up that shows all the info you will need. Usage and Temps for cpu and gpu, framerate, average lows, etc.

Also you might take into account that different games perform differently in terms of what cpu you playing on and what gpu.

The rtx 3070 has been a great card for me with my 1800x. I saw double frame rate increases across most games when comparing it to my 1070ti.

The best way to truly test, is a fresh install of windows and the game, nothing else. And test. Also you might try older versions of the Nvidia drivers.

How was your warzone performance if Ic an ask? With many other games, I have seen significant improvements no doubt, but I still feel like something isn't right.

 

[dainluke]

How was your warzone performance if Ic an ask? With many other games, I have seen significant improvements no doubt, but I still feel like something isn't right.

Just keep this in mind. If you were actually sold an RTX 3070 with 46 SMs and '5888' CUDA cores, then as long as your card is actually boosting to +- 2000 MHz, the G6 is running at 7000MHz (14Gbps) without an OC, and your drivers aren't the issue; then you are facing a bottleneck elsewhere. I kinda hate how there has been this consensus that RAM isn't important and that you just need 16GB of dual-channel RAM. I have literally seen my GPU go from 75% utilization to 100% in Jedi Fallen Order when I upgraded my kit of cheap 2666MHz Vengeance LPX and thereafter did my first ever RAM OC to 3733 C16 with the kit of Dom Plats I had upgraded to, in March. I cannot state how important RAM speed and latency are when trying to get your CPU to adequately draw frames for your high-powered GPU at 1080p. I'm at the point at which I can't in good faith advise friends and fellow competitive gamers to purchase anything less than 3600C18 or 3466C16. 3200C16 is borderline these days tbh. The same goes with HDDs, thus far two of my only friends who own them reported game-ruining stutters.

Rant aside: I think you should just make sure all those technical numbers should be accounted for in RTSS's in-game overlay and perhaps try the DDU. Thereafter you can consider reinstalling COD or perhaps Windows. I really do think it's a simply RAM speed issue though. The fact that more FPS constrained high-demand titles doesn't see this same issue is a clear sign of it to me. At those lower framerates, the amount of data your CPU needs to read and write to RAM to draw the frames simply isn't as high as when your GPU is capable of spitting out 180 FPS and your CPU is being asked to draw 180 frames with slower RAM. Also, when you clocked your RAM to 2666MHz, what did you set your timings to?

 

[∀Z‾soʇɐuɐɥ┴]

So did some testing on my setup.
Max settings at 1080p (incl rtx on), with vsync on I hit a solid 144 fps (obviously), and with it off I run between 180 and 190 fps.
So either your gpu is a lemon, your ram is causing issues, or your cpu is bottlenecking.
CPU bottlenecking seems unlikely unless it's overheating and thermal throttling, so problem is most like with your ram (but I don't see that causing this much of a discrepency in fps, but anything is possible in the wonderful world of computers).

 

[∀Z‾soʇɐuɐɥ┴]

Swapping gpu to the 9700k setup now to see what the fps there is like.

 

[CLAM]

Swapping gpu to the 9700k setup now to see what the fps there is like.

Thanks so much, man, super keen to hear the result.

 

[CLAM]

So did some testing on my setup.
Max settings at 1080p (incl rtx on), with vsync on I hit a solid 144 fps (obviously), and with it off I run between 180 and 190 fps.
So either your gpu is a lemon, your ram is causing issues, or your cpu is bottlenecking.
CPU bottlenecking seems unlikely unless it's overheating and thermal throttling, so problem is most like with your ram (but I don't see that causing this much of a discrepency in fps, but anything is possible in the wonderful world of computers).

Yeah man, it's so odd. Like Timespy and Port Royal score perfectly but Shadow Of Tomb Raider is benching at like 50% of what it should. I think I must just hit up the reseller and ask if we can pop a different 3070 into my rig to test. Let see what the 3070 result is on your side but it may be time to take the card back.

 

[Greekking]

Might be a long shot but check your settings in Warzone, maybe you set a limit to 120fps ?

Remove limits ?

Else maybe give me the 3070 and take my RX480, its a great swap [emoji14]

Sent from my SM-G960F using Tapatalk

 

[∀Z‾soʇɐuɐɥ┴]

Thanks so much, man, super keen to hear the result.

tested on the 9700k with ram running at 3200mhz and got a solid 144fps.
For some reason though even with vsync disabled (in game and in nvidia control panel) it still ran as if vsync was on.

granted, again this is on max settings with rtx enabled etc.

 

[CLAM]

Ok so we definitely know something is not right on my end. What clock speed is his ram at if I can ask and also what did GPU and CPU utilization look like?

 

[CLAM]

Ok so we definitely know something is not right on my end. What clock speed is his ram at if I can ask and also what did GPU and CPU utilization look like?

Wait nvm just saw you said 3200mhz

 

[∀Z‾soʇɐuɐɥ┴]

Ok so we definitely know something is not right on my end. What clock speed is his ram at if I can ask and also what did GPU and CPU utilization look like?

wasn't checking utilization on his end tbh.

 

[CLAM]

wasn't checking utilization on his end tbh.

All good, thanks man, really appreciate the assistance here.

 

[CLAM]

So quick update.
On Shadow of The Tomb Raider, I had anti aliasing SMMx4 enabled which made a 20 frame difference on the benchmark result. With just TAA on I got a few benches between 125 and 130 FPS on the highest preset at 1080 RT disabled.
What I've also found is that turning up resolution pushes the GPU harder but not so much the CPU meaning framerates stay relatively unchanged when pushing up the resolution and other graphics settings. This would indicate a Ram or CPU bottleneck, right? Thing is the difference in a warzone is so massive it still doesn't really explain it. I'm going to go to a mate's hosue and test the card in his rig with a better motherboard and ram, lets see what happens.

 

[ModerndayCowboy]

Since when does a 10600k outperform a 9900k?

Exactly..

 

[ModerndayCowboy]

Just keep this in mind. If you were actually sold an RTX 3070 with 46 SMs and '5888' CUDA cores, then as long as your card is actually boosting to +- 2000 MHz, the G6 is running at 7000MHz (14Gbps) without an OC, and your drivers aren't the issue; then you are facing a bottleneck elsewhere. I kinda hate how there has been this consensus that RAM isn't important and that you just need 16GB of dual-channel RAM. I have literally seen my GPU go from 75% utilization to 100% in Jedi Fallen Order when I upgraded my kit of cheap 2666MHz Vengeance LPX and thereafter did my first ever RAM OC to 3733 C16 with the kit of Dom Plats I had upgraded to, in March. I cannot state how important RAM speed and latency are when trying to get your CPU to adequately draw frames for your high-powered GPU at 1080p. I'm at the point at which I can't in good faith advise friends and fellow competitive gamers to purchase anything less than 3600C18 or 3466C16. 3200C16 is borderline these days tbh. The same goes with HDDs, thus far two of my only friends who own them reported game-ruining stutters.

Rant aside: I think you should just make sure all those technical numbers should be accounted for in RTSS's in-game overlay and perhaps try the DDU. Thereafter you can consider reinstalling COD or perhaps Windows. I really do think it's a simply RAM speed issue though. The fact that more FPS constrained high-demand titles doesn't see this same issue is a clear sign of it to me. At those lower framerates, the amount of data your CPU needs to read and write to RAM to draw the frames simply isn't as high as when your GPU is capable of spitting out 180 FPS and your CPU is being asked to draw 180 frames with slower RAM. Also, when you clocked your RAM to 2666MHz, what did you set your timings to?

3200mhz ram is more than adequate, after 3600mhz there is little benefits to be gained. Most of what you say is true, but 3200mhz should be more than needed for gaming nowadays. Look at min 7:12 in the video linked and there is shown that there is not a big difference in fps. Hence why I said I think it's cpu bottleneck, or could be the mobo with that low ram speed