Random performance drops?

levivh1997

Prominent
Sep 2, 2017
32
0
530
I posted on here before with the same issue, it was solved for a while but now I'm back to step 1. Something I forgot to mention in my first post: this pc has never really worked optimal as about a month after I bought it, it started having random crashes and BSOD. This has never been fixed but went away on its own (?????) only to reoccur every few months for about a week or 2, then go away again.

My specs are:
Windows 10 Pro x64
AMD FX 8320 eight-core processor, 3500MHz
Nvidia GTX 960 (latest update)
2x 8GB Vengeance LP DDR3 DRAM 1600MHz
ASRock 970 Pro3 R2.0
And I play on a 1920x1080, 60MHz screen and a 1366x768 60MHz
All of my drivers are up to date

So the issue is that while playing basically any game I own my FPS and GPU performance drops significantly. I have no issues playing Rainbow Six Siege, DOOM, Star Wars Battlefront. I play all of these games on ultra graphics and changing the graphics quality changes nothing to FPS or GPU performance. I tried benchmarking and I get an average of 52FPS, while benchmarking my GPU is mostly 99% and my CPU is around 80%.

Temperatures are fine, there's no throttle, I never overclocked my CPU or GPU or changed anything in my BIOS. I changed my PSU from an AeroCool Xpredator 750M (80plus bronze) to a Corsair CS650M since I was told that could be the issue.

Changing my PSU did solve the issue for about a week, fps was fine, GPU and CPU usage was fine. Then one day my GPU just went back to its 30% and the fps dropped along with the usage. I restarted my pc, didn't work. I changed the outlet and it worked for about a day or 2. I have tried different outlets to see if maybe that was the issue but no result. A friend of mine thinks it could be the motherboard that was faulty from the beginning since we tried everything else.
 
Solution
Looking back through this thread, you were on the right track with the VRMs however I agree with your friend, i think that board was shot from the beginning. I've seen this particular motherboard you have have problems like this with 125w CPUs quite a few times here. Its just a crap board.

Sorry I know its bad for your budget but you need an 8+2 board.

Vic 40

Titan
Ambassador
Might well be the motherboard. Looking at the vrm's which are without cooling and probably not up to par to give that 8core cpu the power it needs so it throttles or gives the problems you have.
Make sure the ram works at the right timings,speed and voltage as well.
 

levivh1997

Prominent
Sep 2, 2017
32
0
530


the ram is working at the right timings (7-7-7-21), I changed it in the BIOS not too long ago. I found someone online who told me this were the right timings for my motherboard. I can't really afford to buy a new motherboard right now, is there any way to fix this?
 

Vic 40

Titan
Ambassador
You can try and se at what voltage the cpu is stable at current speed or even reduce speed at even lower voltage. See if that will help. Some extra ventilation might work as well.
I'm still assuming it is the vrm right now.So the above may not fix it.
 

Vic 40

Titan
Ambassador

The VRM

MDqtyZU.png

no cooling on it and not very great being just 4+1.For an 8 core cpu you'd like to have 8+2 as minimum with cooling.
 

levivh1997

Prominent
Sep 2, 2017
32
0
530


I put a fan on the VRM but it's a temporary solution. What's the best way to cool the VRM? I was thinking about buying another fan to blow directly on it

 

Vic 40

Titan
Ambassador
http://www.overclock.net/t/946407/amd-motherboards-vrm-info-database
in this link some info about small heatsinks that can be used fo the VRM. Look under "Note 1". Well these are examples of small enough heatsinks that can be used.

You use the stock cooler? Try to use a "top down cooler" which blows air downwards on the motherboard.
Coolers like these,
https://pcpartpicker.com/products/compare/9x3RsY,xzqdnQ,9FGj4D/
 

levivh1997

Prominent
Sep 2, 2017
32
0
530
So I added copper heatsinks to my VRM and I added a fan that now blows directly onto my VRM but my GPU usage is still at 10-30% while playing games like blackwake. I'm starting to think this isn't a temp problem but I don't know for sure.

Apologies for the late response, the heatsinks and fan took a while to arrive.
I've also had more BSOD, mainly about drivers not functioning properly. Most of the BSOD had these errors: kmode_exception_not_handled, kernel_security_check_failure, driver_irql_not_less_or_equal, irql_not_less_or_equal, interrupt_exception_not_handled
 

Vic 40

Titan
Ambassador
Thought about psu,but you already changed that.

Is the ram running at rated specs? Means speed+timings+voltage.Can check with cpu-z (freeware),for current speed and timiings look at the "memory tab" for rated speed and timings look at the "spd tab" at the XMP profile.

Can also make a log,

How to make a log,

download hwinfo32,
install and open it,
check "sensors-only",
click "run",
at the bottom of the window click "logging start",
choose a name and place for the log like "hwinfolog" at the "desktop",
this log can be viewed/opened with either excel from Microsoft or libre office calc.
you can upload it onto a site which facilitates this like SaberCatHost or DropBox,these i like since not so much (or any) spam comes with these.
Just make one until this happens like you described.

You can also use the next to look if it says something about what might cause the bluescreens,
http://www.resplendence.com/downloads
Download WhoCrashed.Lets see if it says something usefull.Use the "Analyze" button and further instructions to see what is going on.Can see if you can copy/paste findings here or make a picture and post that.
 

levivh1997

Prominent
Sep 2, 2017
32
0
530
I downloaded all the things you mentioned, I'm currently running a log of playing blackwake but the fps and usage is all good.
I took some screenshots of the memory and speed tab of cpu-z and will upload them in a bit
 

levivh1997

Prominent
Sep 2, 2017
32
0
530
This is the memory tab: https://imgur.com/UeMACzZ
And the speed tab: https://imgur.com/YTxrU7e both my RAM are the same, same speeds and everything

This is the log file while playing blackwake at around 50-60fps https://www.dropbox.com/s/llb49ctu0d70023/hwinfolog.CSV?dl=0
Since yesterday evening everything has been going fine, no fps drops or stutter, good usage and no more crashes but I don't dare call it fixed yet.
If you want I can record another log while playing total war: warhammer or battlefield 1, those games are mainly the issue
 

Vic 40

Titan
Ambassador
Can't really see voltages,it can be that a motherboard (and AMD more often) isn't fully supported for this tool. So can't be sure about voltages,at first glance are the rest of the values okay.

Migth want to try the other games that are more likely to have issues.

What about the bluescreens? There's also "Bluescreen View" that you could use for info about those.
 

levivh1997

Prominent
Sep 2, 2017
32
0
530
I just tried other games and it's all running fine so far, when it starts getting weird again I'll record a log and link it here
WhoCrashed gave me these messages:

On Fri 9/22/2017 10:45:54 PM your computer crashed
crash dump file: C:\WINDOWS\Minidump\092217-24750-01.dmp
This was probably caused by the following module: ntoskrnl.exe (nt+0x16C580)
Bugcheck code: 0x1E (0xFFFFFFFFC0000005, 0xFFFFF8035AA8C2C0, 0x0, 0x100000020)
Error: KMODE_EXCEPTION_NOT_HANDLED
file path: C:\WINDOWS\system32\ntoskrnl.exe
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: NT Kernel & System
Bug check description: This indicates that a kernel-mode program generated an exception which the error handler did not catch.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.
crash dump file: C:\WINDOWS\memory.dmp
This was probably caused by the following module: ntkrnlmp.exe (nt!KeBugCheckEx+0x0)
Bugcheck code: 0x1E (0xFFFFFFFFC0000005, 0xFFFFF8035AA8C2C0, 0x0, 0x100000020)
Error: KMODE_EXCEPTION_NOT_HANDLED

On Thu 9/21/2017 8:04:09 PM your computer crashed
crash dump file: C:\WINDOWS\Minidump\092117-24093-01.dmp
This was probably caused by the following module: ntoskrnl.exe (nt+0x16C580)
Bugcheck code: 0x139 (0x3, 0xFFFF8881D7B221D0, 0xFFFF8881D7B22128, 0x0)
Error: KERNEL_SECURITY_CHECK_FAILURE
file path: C:\WINDOWS\system32\ntoskrnl.exe
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: NT Kernel & System
Bug check description: The kernel has detected the corruption of a critical data structure.
The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.

On Mon 9/18/2017 12:57:14 PM your computer crashed
crash dump file: C:\WINDOWS\Minidump\091817-24093-01.dmp
This was probably caused by the following module: ntoskrnl.exe (nt+0x72127B)
Bugcheck code: 0x124 (0x0, 0xFFFFD4872E3288F8, 0x0, 0x0)
Error: WHEA_UNCORRECTABLE_ERROR
file path: C:\WINDOWS\system32\ntoskrnl.exe
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: NT Kernel & System
Bug check description: This bug check indicates that a fatal hardware error has occurred. This bug check uses the error data that is provided by the Windows Hardware Error Architecture (WHEA).
This is likely to be caused by a hardware problem. This problem might also be caused because of overheating (thermal issue).
The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.

On Mon 9/11/2017 8:47:14 PM your computer crashed
crash dump file: C:\WINDOWS\Minidump\091117-23781-01.dmp
This was probably caused by the following module: dxgmms2.sys (dxgmms2+0x63B7)
Bugcheck code: 0x3B (0xC0000005, 0xFFFFF80E1E8063B7, 0xFFFFAA8014D22530, 0x0)
Error: SYSTEM_SERVICE_EXCEPTION
file path: C:\WINDOWS\system32\drivers\dxgmms2.sys
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: DirectX Graphics MMS
Bug check description: This indicates that an exception happened while executing a routine that transitions from non-privileged code to privileged code.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
The crash took place in a standard Microsoft module. Your system configuration may be incorrect. Possibly this problem is caused by another driver on your system that cannot be identified at this time.
 

levivh1997

Prominent
Sep 2, 2017
32
0
530
Pc just crashed again while just browsing youtube, it's been running fine for the entire day, even after the crash everything is fine. There was no BSOD just this sound and the pc didn't respond anymore.
 

Vic 40

Titan
Ambassador
Looked abit at the bluescreens and i think the first three are something that is that a problem occurs and since not knowing exactly what it is something else in the windows kernel get's the blame but really isn't the problem.
The last seems to point at DirectX,maybe the gpu has issues,maybe the drivers don't work well with it.
Can you test with another gpu? Maybe from a friend or local shop ...
Can you try not the latest drivers,but an older set,try some revisions before like say 5.
https://www.geforce.com/drivers/results/117918 (=the latest 381.xx set)
Can clean these latest you use with DDU or when installing the new drivers you want to test use advanced mode and check the "Clean Install" option.
If more bsods happen wil i ask for some help with those. There's someone here that is just better with handling those.
 

levivh1997

Prominent
Sep 2, 2017
32
0
530
I already tried setting the driver to an older version but it didn't do anything. I can't really change or replace the gpu now but I'll look into it if someone I know has a spare gpu they don't use
I also made a new log from when the fps and usage was acting up again, maybe you can see something in this log: https://www.dropbox.com/s/5z0djfb4y4wjx6x/log%204.CSV?dl=0
When I shut down the pc and restarted it immediatly crashed before even properly loading in the background, programs, etc
 

Vic 40

Titan
Ambassador
Can you look with "AMD Overdrive" at the temps of the cpu? This tool has a better way of looking at the temps.Other tools just don't seem to look at them properly/read the sensors well. The bigger the "Thermal Margin" the better. In the red is not good.

Sorry to ak this so late,could have thought of that before,but in the logs seem temps fine,but cannot be sure. Cannot really find something to blame in the logs. Cannot look at the +5V and +12V rail so can't see if these give issues. You might be able to check these in the bios,but that will be values without load.If already close to spec limits could these be a problem too.

Sudden low fps and the crash right after rebooting seem to indicate a hardware failure which i cannot truly pinpoint. I do see that the cpu doesn't go to full turbo (4ghz) so is maybe hampered by temp limits. In that case is also maybe cpu temps an issue to look at.
I also see that the cpu goes to 100% at one time where the gpu does even drop to 22% load so it looks like a cpu bottleneck,but don't see why that cpu should be bottlenecking this gpu.Maybe another idication of temp issues. If this is the cpu itself or still the VRM is hard to say.

What cooler do you use on the cpu? Does it help if you disable the turbo option in the bios?

 

levivh1997

Prominent
Sep 2, 2017
32
0
530
I disabled the cpu turbo in the bios, I'm still to test this out as my pc is now installing AMD overdrive

I replaced the stock cooler with a cooler master hyper 212 evo, I've had this cooler for over a year now since the guys at a pc shop told me to replace the stock cooler with a better one, they also installed it so I'm pretty sure it's installed correctly
 

levivh1997

Prominent
Sep 2, 2017
32
0
530
I've also thought about a cpu bottleneck and did a bit of researched but some people say the cpu isn't good enough for my gpu and others say there shouldn't be a problem between my cpu and gpu so I don't really know what to make of it.
 

levivh1997

Prominent
Sep 2, 2017
32
0
530
So I recorded my cpu while playing Battlefield 1: https://www.youtube.com/watch?v=sD9_mDT2zNs
I apologize for the bad quality, I tried to line up my phone with the screen to get the best quality
It went well for about 30 seconds, after that I had stutter, fps dropped to 14-20 so I recorded for a bit and then shut it off.

This is right after Battlefield 1 when I tried Blackwake: https://www.youtube.com/watch?v=UiBRnZDo8hs
Now I got 50-54fps, no stutter, no nothing. I also put the gpu graph from msi afterburner next to it (I forgot to do this with Battlefield 1) to kind of get a perspective to both of them.

This is what AMD Overdrive shows for my cpu: https://imgur.com/2v77aBP
It doesn't show my gpu for some reason, everything is blank when I click the gpu tab
Board tab: https://imgur.com/Ry16sfX
Clock/Voltage tab: https://imgur.com/ZKo9kam
 

Vic 40

Titan
Ambassador

Looked at that video,i guess that youplayed the whole time? What i see is that at some point even 6 cores drop to lowest frequency possible which is why imo you have such bad fps,

26Qoe2r.png

I know also see that temps in the thermal margin are rising which should mean that it get's cooler,but wonder if it really get's hotter.The next says that "Thermal Margin" is the difference between the what the cpu can handle as max temp and the current temp,
https://www.google.nl/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwjjrLrH1sXWAhXJJ1AKHSuIAaUQFggmMAA&url=http%3A%2F%2Fwww.tomshardware.com%2Ffaq%2Fid-2122665%2Funderstanding-temperature-amd-cpus-apus.html&usg=AFQjCNEi0g7DyNQTSv_1XYmYDjnrMGSKig
but wonder if that is a setting you can change in Overdrive.Maybe it can really tell core temps as well.
However if it is as told above does that mean that the cpu actually get's cooler and so must it use less energy/Vcore which might lead us back to the motherboard again.
There's even a point where all cores are at minimum frequency at min. 1.51.I don't see these kind of drops in voltage (vid) and frequency in the second video although temps are quite the same,well not going above 49°C and more steady..

I didn't see the video so can't know when maps are changed or this a ongoing game so if drops are related to that.

Going to throw this one in the group as well and ask others for some advice as well.See if they can make my assumption fail.

Does it btw still (when playing BF1) give drops,and the fps loss with that, when disabling "Cool'nQuiet" in the bios?
 

levivh1997

Prominent
Sep 2, 2017
32
0
530


I joined a quickmatch and it was one of the newer maps, I played for a total of 3 minutes maybe, the very beginnig few seconds are when I started the game up, in the menu it ALWAYS has 60+ fps and 99% usage (both cpu and gpu) and as soon as I start a game it drops

I'll test it out but it's hard since the drops don't always happen, it's more of a bad luck situation at this point. I "hope" the drops happen this evening so I can test it out. Btw my cpu turbo was set to auto in the bios, I don't know if this might mean anything?

And thanks a whole lot for your help so far! I really hope it won't be too complicated. So far even the people at pc shops haven't been able to figure it out
 

Vic 40

Titan
Ambassador
Try disabling Cool'nQuiet and see what that gives.
Also as second test set cpu Turbo to disabled.

Try these suggestions first each on their own,and maybe after that combine.lets see what results these tests give.

And you're welcome.:) Have asked some others to look at it too in another thread.