Identifying a bottleneck. Help Needed!!

[NaysWindu]

Hi Everyone. This is gonna be a big post

 

TLDR: Weirdass stutting issues caused by neither the CPU nor GPU

 

I have been experiencing various fps/hardware issues for the last couple of years. Normally, I chalked it up to my OS being bloated, or my GTX 970's terrible chip (lost hard in the silicon lottery) - 

  (not using furmark anymore, due to its issues)

Basically I couldn't apply an OC of more than 25 Mhz before I had stability issues, and I couldn't add more than about 3mV (I still have more to learn about overclocking so some more advice here would be nice )

 

Anyway, to the point of this post

Recently, I was able to get my hands on both a 1080ti, and an Acer Predator z271, for a very good price. Once I had entered the realm of high refresh rate gaming, I began to notice some issues. 

1 - Framerates in Esports games (specificially League of Legends), would get barely above 144 fps (vsync off)

2 - Framerates in AAA games were particularly bad in certain areas. This was most obvious and easiest to reproduce was in Destiny 2, where when loading into the Titan patrol zone, I would immediately hit 50 fps, and struggle to get above 60 when viewing the patrol zone. Looking away into the skybox would restore the framerate (but it would still be lower than expected).

- 3dmark firestrike seems a little poor on the shared load score, but I am not sure about this one.

- Issues are also observed in other games such as PUBG, Apex Legends, and especially in Battlefield 5 (mega stuttering).

 

Now at different points, I've been able to get replacement parts for almost everything in my system, and have tried every possible configuration

I have also done multiple fresh installs of windows 10, and have used DDU several times when changing out GPUs. GPUs are always in either the adaptive, or max performance state through the Nvidea control panel.

Slashes have been used when there have been 2 components available. I've abbreviated names where possible, but the experience members in the community should know the parts I am talking about

 

CPU - Intel i7 4770 /  Intel i5 4690 (dont currently have)

GPU - EVGA GTX 970 FTW (first model) / ASUS strix 1080ti Gaming

Motherboard - Gigabyte Z87-HD3 / Asrock Z87-Extreme4

RAM - Corsair Vengeance DDR3 @2400Mhz 2x8GB / Strontium SIRT4G86U1 DDR3 @1333Mhz 2x4GB

Drives - Corsar Force LS 240GB SSD, WDGreen 480GB SSD, Seagate Barricude 2TB HDD (not plugged in right now)

CPU Cooler - Intel Stock Cooler/ CoolerMaster Hyper 212X

PSU - Corsair AX860

 

I have attatched a HWinfo page on the current setup. Only thing that isn't so good is the RAM, but the issue will persist even with the corsair RAM back in.

 

 

I have also changed case in all of this, but most tests have happened with the side panels removed.

Everything left of the slash was in the original configuration when the issue was first noticed (as far back as 1.5 years ago)

Firestrike tests:

GTX 970

https://www.3dmark.com/fs/20035300

https://www.3dmark.com/fs/20032812

 

GTX 1080ti

https://www.3dmark.com/fs/20032380

https://www.3dmark.com/fs/19948252

 

All tests have been attempted all of the possible hardware configurations from the list above. I have attempted running windows of different drives, and even running them by themselves.

Issue 2 occurs regardless of if the 970 or the 1080ti is being used. Similar framerates with both at the same game settings. TEMPS ARE COMPLETELY UNDER CONTROL. Neither CPU not GPU have risen above 70C for any of this.

ALSO - Neither CPU Nor GPU are at 100% in either of the tests (except 3dmark, but in the Shared test they were both chilling around 70% usage)

I have also taken the boot SSD with Destiny on it, and run it on my dad's computer, with the same destiny 2 test. His computer had no issues, and was rock solid with the framerates (didnt dip below 60 (had v-sync on) (my issue occurs regardless of v-sync on or off))

 

His specs - 

CPU - Intel i7 6700k

GPU - EVGA GTX 970 SC

RAM - Corsair Vengeance LPX DDR4 2x8GB (not sure what its clocked at)

M'Board - Gigabyte z-170 Gaming 7

Drive - just the WD green drive from my system

PSU - EVGA Supernova 850G2

CPU cooler - Some tower cooler with 2 fans. Looks beefy enough

 

 

I've just about given up. The only option that I have left is to try out my dad's PSU in my system, to see if that fixes anything. I doubt it will make a difference, but its the only component that hasn't changed that was part of the original system.

All values and temps were recorded using MSI afterburner. If there is demand, I can run the tests again, and record the logs or something, and include them in this post. I am coming to you guys, because this is the most amazing tech community out there, and I've seen so many people get help when they needed it. I'm gonna be out for a couple of hours from when I post this, but when I get back, I'll start doing thing suggested, if there are any. Sorry for the long post.

If there are any bits of information that need to be added, let me know, and I'll add them

 

Love you all.

 

 

DESKTOP-C3I1BCM.HTM

Edited August 21, 2019 by NaysWindu
Added HWinfo page. Updated 2nd comp PSU details

[TheGreatSora]

try out a different PSU. Also try to run it with the corsiar memory only 

[Turtle Rig]

19 minutes ago, TheGreatSora said:

try out a different PSU. Also try to run it with the corsiar memory only 

Also try to run DDU "Display Driver Uninstaller" in safe mode and do a clean and restart.  Then install the latest WHQL drivers and let us know what happens.

[NaysWindu]

2 hours ago, TheGreatSora said:

try out a different PSU. Also try to run it with the corsiar memory only 

Issue first occurred when i was running with only the corsair RAM. Only got the new set and tested it today, before posting.

edit: also getting my hands on the other PSU might be a few days, as that computer is being used. An AX860 should be fine though right? PSU's dont normally cause these kind of issues?

Edited August 19, 2019 by NaysWindu

[NaysWindu]

2 hours ago, Turtle Rig said:

Also try to run DDU "Display Driver Uninstaller" in safe mode and do a clean and restart.  Then install the latest WHQL drivers and let us know what happens.

I did multiple driver re-installs using DDU. I'm assuming all Nvidea game ready drivers are WHQL signed? I even tried older drivers, with no discernible difference

[Turtle Rig]

1 hour ago, NaysWindu said:

Issue first occurred when i was running with only the corsair RAM. Only got the new set and tested it today, before posting.

edit: also getting my hands on the other PSU might be a few days, as that computer is being used. An AX860 should be fine though right? PSU's dont normally cause these kind of issues?

500 watts is even good.  What is important is the ampage on the 12v rail.  Some PSUs have more juice then others.  Go by price and you will see what I mean.  A decent 550w would be good enough for a future video card upgrade and what not.

[NaysWindu]

So update: I recently got another monitor from a friend. I set it up and ran the same destiny test on it. I got a noticeable performance improvement (roughly 20-30 fps), but still not enough to account for the problem. My acer monitor is currently at acer support for a defective panel (well documented issue with the z271, i got it for super cheap so i didnt really mind)

V-sync was off for both tests

3 monitors total

acer z271- at support

 

previous monitor - Dell s2409W (1920x1080) that I overclocked to 75Hz. Uses an HDMI connection.

new monitor - Dell U2312M (1920x1200) that is overclocked to 62Hz. Uses display port. Couldn't push it any further :/

 

Maybe the connection standard makes a difference? That would suggest something power related? I don't really know

I'll be testing the second power supply tonight. Hopefully it will show a difference

[NaysWindu]

On 8/20/2019 at 12:19 AM, Turtle Rig said:

500 watts is even good.  What is important is the ampage on the 12v rail.  Some PSUs have more juice then others.  Go by price and you will see what I mean.  A decent 550w would be good enough for a future video card upgrade and what not.

Ok so I've tried my dads PSU now (EVGA Supernova 850G2) and there was no change. At this point I'm figuring that i should just sell everything for parts, and go for an x470+ryzen 3rd gen upgrade. Don't know if that would fix it tho  Maybe I'm cursed? The only other possibility I can think of is that I have 2 bad motherboards. But that isn't likely. I also know that it isn't the gpus, as I've tried my dad's 970, and the issue persisted, but when booting on his system, it didnt.

[GazJ]

i used to have the 960 ftw and i know that would boost to around 1600mhz

have you set the power plan in nvidia cpanel to performance? and try clicking k boost in evga software

[NaysWindu]

12 hours ago, GazJ said:

i used to have the 960 ftw and i know that would boost to around 1600mhz

have you set the power plan in nvidia cpanel to performance? and try clicking k boost in evga software

I set everything to max performanance in the cpanel. Currently have the 1080ti in.

[NaysWindu]

Ok. at this point, ive given up. Ill probably just sell everything i can and upgrade to ryzen 3000. Thanks to those that replied :)

[Waqas409]

Use the nvidia driver version of 3xx. Don't go into 4xx ones they will give you poor performance no matter how good your setup is