[SOLVED] Extremely LOW FPS while gaming 3080ti STUMPED!

[varsxusar]

Good Morning Everyone,

I am 1000% stumped. I finished a new build a couple of weeks ago and everything was fine and dandy. Fastest system ive ever built. Until a couple of days ago when issues came out of left field... no beyond left field.
When I would boot up my pc, it would go through these weird power cycles of turning off and on several times and eventually I would get to the windows log in screen. Other times I would boot up and once I logged in, a couple of minutes later everything would freeze however I could still hear audio (like friends talking over discord). After these odd events and rebooting it would be like nothing ever happened. I would load up and game for hours no issues. I tried several things like reconnecting everything to make sure that it was all connected properly (including gpu, ram sticks, etc) I even made sure all drivers were up to date and reinstalled drivers.

Now the issue has changed.... I seem to boot up just fine and have not frozen yet. But when I run any gaming application I am getting 8-15 FPS. I launched Hunt Showdown, Escape from Tarkov and even tried Heaven Benchmark and each time only 8-15 FPS. The oddest part is videos on netflix, youtube, etc are all 100% fine. After a couple of restarts and trying a new display port slot it was randomly back to normal, but then again this morning after I booted up, 8-15 fps..... All this to say that I have zero clue what went wrong or whats going on.

Specs:
5950x AMD @ 4.3ghz (was 4.5ghz but tried lowering in case that was the issue)
EVGA 3080ti FTW3
64gb Corsair Vengeance RGB PRO 3600 (4 sticks)
1TB WD_BLACK NVMe SSD
SilverStone 1000W PSU
Lian Li Vertical GPU mount (PCIe 4.0)

Thanks to everyone and anyone who took the time to read this shhhh show

 

[Frag Maniac]

This didn't by chance start happening after you began using Discord with the new PC did it? I've read Discord can often cause problems with games, including lots of crashing. Maybe try uninstalling the app if you have it on your PC, then do a malware and system file corruption check with Malwarebytes and SFC.

I would also say check all hardware connections, but if it initially ran fine, and you haven't added any hardware, that may be moot. Did you initially test it hard with graphics intense games though? If you have access to a spare GPU and PSU you can test, perhaps via a friend, that might help narrow it down.

Also make sure your RAM is running at it's rated speed by setting it to XMP mode. This is crucial with Ryzen CPUs, as their "Infinity Fabric" interconnect structure is reliant on good RAM speed to run properly. It wouldn't hurt to test with one or two sticks of RAM at a time either.

Most don't run 4 sticks of RAM in a gaming PC, and with Ryzen's Infinity Fabric causing much higher latency than Intel's CPUs, I'm not even sure it's a good idea, as latency can also be a problem when gaming on 4 sticks of RAM.

Malwarebytes
https://www.malwarebytes.com/premium

SFC - Follow instructions exactly
https://support.microsoft.com/en-us...-windows-365e0031-36b1-6031-f804-8fd86e0ef4ca

 

[varsxusar]

I dont think discord is the problem as I would normaly play Escape from Tarkov while streaming it onto discord and watching my other 2 friends stream EFT as well all at the same time no issues (Game on Monitor 1, Discord on Monitor 2). Still I tried uninstalling discord but the issue persisted.

All the connections look to be good to go, I dont have a spare psu/gpu at this time but I dont suspect this to be the issue. I did run benchmarking testing the GPU/PC when I first built the pc and everything was exceptional, and all games ran extremely well with high fps no stutters or spikes.

I do have my ram set to its correct speed with XMP (or AMDs version of it) enabled, Also set at the correct voltage. I also tested the ram using the windows diagnostic tool and came up with zero errors. I will try running with only 2 sticks of ram (32gb 2x16) and see if theres any changes but im not too hopefull.

I ran malwarebytes and came up with a couple PUPs that were taken care of as well as the SFC and that also came back to 100% verification with no issues. Unfortunately no luck, issue still persist.

im think im boned! lol this is just so odd that it would be fine one day and then poof issues.

 

[Zerk2012]

I dont think discord is the problem as I would normaly play Escape from Tarkov while streaming it onto discord and watching my other 2 friends stream EFT as well all at the same time no issues (Game on Monitor 1, Discord on Monitor 2). Still I tried uninstalling discord but the issue persisted.

All the connections look to be good to go, I dont have a spare psu/gpu at this time but I dont suspect this to be the issue. I did run benchmarking testing the GPU/PC when I first built the pc and everything was exceptional, and all games ran extremely well with high fps no stutters or spikes.

I do have my ram set to its correct speed with XMP (or AMDs version of it) enabled, Also set at the correct voltage. I also tested the ram using the windows diagnostic tool and came up with zero errors. I will try running with only 2 sticks of ram (32gb 2x16) and see if theres any changes but im not too hopefull.

I ran malwarebytes and came up with a couple PUPs that were taken care of as well as the SFC and that also came back to 100% verification with no issues. Unfortunately no luck, issue still persist.

im think im boned! lol this is just so odd that it would be fine one day and then poof issues.

I would take the riser for the card out and try it straight into the motherboard.

 

[Sohom]

Discord do cause issues for some... anyways, first of all, mount without the pcie extension and Check the ram, i.e, try with just 1 stick or 2 in dual memory channel... and about latency, this gen ryzen has come a long way...

 

[varsxusar]

So I ran some test.... I just noticed that my GPU clock is stuck at 420mhz and Mem clock at 9501mhz.... when they should be boosting to damn near 1800mhz.

I still have to try removing the pcie extension (its a pain in the butt since I have this water loop set up going on the vertical card)

 

[varsxusar]

I dont know what this did if it did anything at all or just plain dumb luck....

I read somewhere that you may want to unplug power cord and hold power button for 30 seconds to reset... idk everything or whatever... so I tried it and now my clock speeds on GPU are up to 1950mhz and games are no longer laggy... what in the hell is going on

 

[Frag Maniac]

I was under the impression the vertical mount GPU was due to the case design receiving the MB the other direction, vs an adapter being used. Otherwise I'd definitely had recommended going without the vertical mount Pci-e adapter. I don't recommend adapters of any kind on MBs, especially on something as critical as the GPU. Just too much risk with connection reliability and latency.

Then again, you said you tested thoroughly with games first and this didn't happen, and I didn't hear you say it wasn't set up that way initially. So I would try to think hard if you made any changes at all, software or hardware wise, before this started happening. Even plugging the PC into a different wall outlet, as sometimes those can be problematic too.

I would also not assume certain parts like the PSU or GPU are rock solid if this persists. Just because you had good results initially doesn't automatically mean a component didn't poop out on you. I mean you can usually see 98% of quality products in a model line do well, but there's always that 2% that isn't up to snuff, and it's like a lottery, as they can't thoroughly test ALL of them. In that regard, we the consumers are sort of their last catch all as far as product testing goes.

 

[varsxusar]

I was under the impressing the vertical mount GPU was due to the case design receiving the MB the other direction, vs an adapter being used. Otherwise I'd definitely had recommended going without the vertical mount Pci-e adapter. I don't recommend adapters of any kind on MBs, especially on something as critical as the GPU. Just too much risk with connection reliability and latency.

Then again, you said you tested thoroughly with games first and this didn't happen, and I didn't hear you say it wasn't set up that way initially. So I would try to think hard if you made any changes at all, software or hardware wise, before this started happening. Even plugging the PC into a different wall outlet, as sometimes those can be problematic too.

yeah it was running just fine before which makes this whole thing super strange. for reference I have the Lian Li (who doesnt) 011 mini with the lian li vertical mount which comes with a PCIe 4.0 extension in order to mount the GPU vertically.

Its just so bizarre and it would be nice to know what the issue is to try and avoid it in the future. As I said before, after I unplugged power and held the button down for a bit everything seems to be back to normal but this has happened before then it suddenly goes bonkers again so we will see

Truly appreciate the suggestions and ill keep this updated with any changes or if I happen to figure out if something specific causes the issue

 

[Frag Maniac]

I dont know what this did if it did anything at all or just plain dumb luck....

I read somewhere that you may want to unplug power cord and hold power button for 30 seconds to reset... idk everything or whatever... so I tried it and now my clock speeds on GPU are up to 1950mhz and games are no longer laggy... what in the hell is going on

Are you sure at some point when you went into the BIOS to enable XMP, you didn't accidentally save Default settings vs adjusted settings? Default settings are like a factory reset to protect components, which sets everything to lower speed. Generally the clocks wouldn't have dropped THAT low while gaming though, just at idle

Could as well just be a fluke with your MB settings needing rebooting though. If so, hopefully that's just a one time thing. Have you updated your MB chipset drivers and looked for any necessary BIOS updates? I mean that CPU is one of the newer high end ones, so it's worth checking to see if the BIOS you're running is fully compatible with it.

...for reference I have the Lian Li (who doesnt) 011 mini ...

I remember when Lian Li used to be THE talk of alu gaming cases. My first custom gaming PC I had built by a local shop using a Kingwin KT-424 alu case, which were pretty popular at the time, and very affordable.

The cooling was handled by three 80mm fans, 2 front, 1 rear, but the fronts had a dust filter. My how things have changed!

https://www.newegg.com/p/N82E16811170012

I've always thought Lian Li was one of the more high quality alu cases, but I can't say I'm impressed with a design that gives the user the option to install a GPU adapter, whether encouraged or not.

 

[varsxusar]

Are you sure at some point when you went into the BIOS to switch XMP on, you didn't accidentally save Default settings vs adjusted settings? Default settings are like a factory reset to protect components, which sets everything to lower speed. Generally the clocks wouldn't have dropped THAT low while gaming though, just at idle

Could as well just be a fluke with your MB settings needing rebooting though. If so, hopefully that's just a one time thing. Have you updated your MB chipset drivers and looked for any necessary BIOS updates? I mean that CPU is one of the newer high end ones, so it's worth checking to see if the BIOS you're running is fully compatible with it.

Well I tried a couple things. I reset My bios to factory default settings then benchmarked, no fix, I then tried a 4.65 OC and made sure ram was set at correct timings/voltage. Still no fix. there was a vbios update for the GPU but after the reset still nothing. Im gonna check right now to see if theres a bios update and if there is ill be doing that asap

 

[Frag Maniac]

Im gonna check right now to see if theres a bios update and if there is ill be doing that asap

Generally I don't bother with BIOS updates unless I'm having a problem that can't be fixed otherwise. Since you're having strange problems and have a new model CPU though, I'd say it's advisable, especially If you've not even checked or read up on any possible BIOS updates. I DO advise reading what they're for before installing them though.

 

[varsxusar]

Generally I don't bother with BIOS updates unless I'm having a problem that can't be fixed otherwise. Since you're having strange problems and have a new model CPU though, I'd say it's advisable, especially If you've not even checked or read up on any possible BIOS updates. I DO advise reading what they're for before installing them though.

well here we go again... once I took a look, I was about 3 bios versions behind (like you I dont bother with updates much on bios)
Just updated the bios and set bios settings to my OC settings as per usual and again, stuck at 420mhz gpu clock. I did notice that when im having this issue, EVGA PX1 shows everything at zero. 0 clock, 0 memory clock, 0 voltage.

When it was working fine, the clock speeds would be jumping around and depending on the application I was running.

 

[varsxusar]

just reinstalled video card drivers and rebooted... and its good to go again... what a cycle well see how long this last but still cant determine what if anything is causing the issue.

 

[conofroc75]

What MoBo do you have? Sometimes the software for controlling things like MSI dragon center or now its Control center can be wonky with profiles and they can set voltages and clocks to things which can cause issues. For example if its a ing mode or not could maybe imposing clock limits or just being weird in general. Its really a pain too because sometimes it van set voltages higher than advisable on certain things. They have been getting much better more recently but its always worth taking a look if your having problems. Also a goof thing to know is if you open up nvidia control panel and look under help and click debug mode that will set the gpu to its base clock and disable anytype of software imposed limits so if it happens again maybe try that so then you can narrow it down a bit more

 

[varsxusar]

What MoBo do you have? Sometimes the software for controlling things like MSI dragon center or now its Control center can be wonky with profiles and they can set voltages and clocks to things which can cause issues. For example if its a ing mode or not could maybe imposing clock limits or just being weird in general. Its really a pain too because sometimes it van set voltages higher than advisable on certain things. They have been getting much better more recently but its always worth taking a look if your having problems. Also a goof thing to know is if you open up nvidia control panel and look under help and click debug mode that will set the gpu to its base clock and disable anytype of software imposed limits so if it happens again maybe try that so then you can narrow it down a bit more

I have a Asus Crosshair VIII Hero

Well im sad to report that after it being ok last night.. I booted up this morning and again.... FPS down to the 10s and GPU clock speed stuck at 420mhz....

😡

 

[varsxusar]

Update:

I was finally able to dig out my standard backplate to mount the GPU directly into MOBO without using the vertical GPU extension.

After booting up the screen momentarily remained black with a small pixel line almost like a screen tear. it then loaded into windows just fine and it is also working fine!

I have ordered a new extension incase the one I got was bad.

The current extension being used is the one that comes with the Lian Li 011 Vertical mount backplate for PCIe 4.0.

Im happy to report that after rebooting, we are still looking good. Before, after being ok and rebooting it would go back to crappy mode. Ill update again later once I get the new PCIe 4.0 extension and vertical mount the GPU.

So far looks like the Lian Li PCIe 4.0 extension may be the culprit.

 

[Frag Maniac]

So far looks like the Lian Li PCIe 4.0 extension may be the culprit.

Like I said, bad case design to include a super long Pci-e vertical mount adapter extension, since they can increase latency and are known to have spotty reliability.

I really don't get why you still insist on using that setup, vs just mounting it horizontally. I know, I know, you said it's due to your rad config, but a case that versatile ought to be able to adapt to different rad placement easily.

I always feel like a horizontally mounted GPU is better. It avoids excess dust settling into the fans and heatsink, plus it draws air from down low, which is typically the cooler air in the case.

Your choice though, hope you get lucky this time.

 

[varsxusar]

Like I said, bad case design to include a super long Pci-e vertical mount adapter extension, since they can increase latency and are known to have spotty reliability.

I really don't get why you still insist on using that setup, vs just mounting it horizontally. I know, I know, you said it's due to your rad config, but a case that versatile ought to be able to adapt to different rad placement easily.

I always feel like a horizontally mounted GPU is better. It avoids excess dust settling into the fans and heatsink, plus it draws air from down low, which is typically the cooler air in the case.

Your choice though, hope you get lucky this time.

The vertical GPU setup was more so part of the custom loop design im going for here shortly. Im not opposed to changing it a bit and mounting the GPU the traditional way.

I just got the new PCIe 4.0 extension and remounted the GPU vertically with the new extension and rebooted. All systems were a go so im thinking a faulty extension might be to blame. Ran heaven benchmark without issues.

Im gonna keep an eye on it rebooting throughout the day and if everything is fine then we found the culprit. If the issue comes back then as you mentioned, ill just have to go with the traditional mounted GPU and change my loop design. Hopefully my next update is a good one.