PC Locks While Gaming Only - PSU Dying?

So starting around 2 or 3 months ago, my pc would randomly lock up while gaming and make a loud buzzing noise. I have to hold down the power button to do anything from that point. It was random at first, I could go for many days without this happening, but it has now gone to a rate of once every 45 minutes or so. Again, this only happens when gaming.

I have been sleuthing the internet and tried many tweaks to settings, reinstalled Win 10 twice, made sure GPU drivers were up to date, etc. Nothing has worked, and I am now wondering if it is hardware related instead given the lackluster success with a software focused approach and that it has become more frequent with time.


i7 7700k (OC'd to 4.8 Ghz)
EVGA GTX 1080 SC
16 Gb Corsair Vengeance LPX DDR4 3000
Win 10 64 Pro
Z270 m7 mobo
Cryorig R1 Ultimate cooler
1Tb Samsung 850 Pro
1Tb Samsung 850 EVO
1TB WD Black Mechanical
Corsair RM 850x Gold PSU

This happens on any and all games, modded or not, Origin, GOG, Steam, etc. Right now I am only trying out MGSV-TPP and that is the only game installed, period.

The only thread that was quasi useful with the same problem mentioned the load on the psu from an sli setup, and it got me wondering if I might have a hardware issue.

Thoughts? I am totally desperate here, I have a very pretty paper weight on my hands as this is a gaming rig only. Normally tooling around the web gives me a fix, but this time I'm coming up empty.

Could this be a psu or other hardware related issue? Anyway to figure out if it is, or what other component could be the culprit?
10 answers Last reply
More about locks gaming psu dying
  1. I doubt that it would be PSU issue since:
    1. You have good quality PSU with more than enough wattage.
    2. The system freezing symptom doesn't point towards PSU. If it were PSU issue, you'd see reboots/shut downs.

    As far as what might be the issue, i'm leaning towards software issue. Still, there are some tests to do.

    To know if any of your component is underperforming, run UserBenchmark test tool,
    link: http://cpu.userbenchmark.com/Software

    Post the results link here for further validation.

    Also, run Unigine Superposition at the highest settings your hardware can take comfortably (i'd say medium settings @ 4K or ultra settings @ 1440p),
    link: https://benchmark.unigine.com/superposition

    If Unigine Superposition does freeze during the run, just like all your games, remove your CPU and GPU OC. Then, run Unigine Superposition again with stock clocks on CPU and GPU.

    Oh, what kind of temps you're seeing out of your CPU and GPU while gaming? Also, what are the idle temps?
  2. First, thank you for taking an interest, I truly appreciate it.

    So UserBenchmark gave me this

    http://www.userbenchmark.com/UserRun/9495343

    And here are the results of the Unigine runs. Shaders had 4k and 8k Optimized options, but I chose not to select those on either test run. I also normally don't run the GPU overclocked since 1) the difference is negligible to nil in games I normally play (I'll use afterburner normally, and that's to set fan speed only) and 2) I really don't know how to safely push the GPU OC and have waited until I need to do it to learn how.

    Unigine first run: MB OC on, XMP enabled - RAM at 3000Mhz, GPU OC Off
    Setting: 4K, Full Screen, Extreme Shaders, High Texture Quality. DoF On, Motion Blur on (VRAM usage at 4442 out of 8192)

    Results-No crash

    FPS Min 4.55, Avg 5.38, Max 6.20
    GPU Temp Min 30.0 C, Max 77.0 C
    GPU Utilization Max 100%


    Uningine second run: MB OC on, XMP enabled - RAM at 3000Mhz, GPU OC Off
    2K, FullScreen, Extreme Shaders, High Texture Quality. DoF On, Motion Blur on (VRAM usage at 3612 out of 8192)

    Results - No Crash

    FPS Min 13.76, Avg 16.00, Max 18.40
    GPU Temp Min 37.0 C, Max 80.0 C
    GPU Utilization Max 100%


    Idle CPU Temps

    Avg 35 C - From HWM


    Gaming/Load CPU Temps - I never check CPU temps much while gaming, but I recall around 45 C-ish from Afterburner OSD when I used to run that, and HWM was showing low to mid 40s when I ran MGSV earlier to double check.

    GPU Temps at idle are normally mid to high 30s and gaming are usually around the 75 C give or take, and can range up to 80 - 82 C. But that is not common and being sure the fan is going is the normal "fix" for that.

    Thanks again. I am being relegated to playing my Xbox one atm (I know, first world problems huh?) while I try and work through this. Still, when you spend 2-3k and the time to go homebuilt, ya kinda sorta wanna use it too :)
  3. Hmmm... your GTX 1080 should ace the 1440p (2K) benchmark in Superposition with 60+ FPS. Select the 1080p Extreme preset in Superposition and look what kind of FPS and temps you'll get out of your GPU.

    You can also go with weaker presets, e.g 1080p Medium and look at which point you'll get your GPU running at 60 FPS in the benchmark.

    Also, your gaming temps for GPU is too high. I'd expect such temps from manual high-level OC'd EVGA GTX 1080 SC and not from the non-OCd GPU.
    What kind of PC case you have and also how many fans with the orientation, size, make and model? Oh, fans RPM too (not the max but the ones you have them running at).
  4. OK, so here are some more data... I ran superposition several times. Here are the numbers

    The first run had similar results to the last time I ran them...


    first run: MB OC on, XMP enabled - RAM at 3000Mhz, GPU OC Off
    2K, FullScreen, Extreme Shaders, High Texture Quality. DoF On, Motion Blur on (VRAM usage at 3470 out of 8192)

    Results - No Crash

    FPS Avg 15.98
    GPU Temp Avg mid to high 70s, max recorded at 81
    Fan Speed/Tach (on auto profile) ~ 51% at ~ mid 1400 rpms
    CPU Temps Avg mid to high 40s (cores 1 & 2 mostly low 40s with 2 spikes around the low 50s; cores 3 & 4 with 2 coinciding spikes around the high 60s - both spikes were only a second long, if that).



    The second run had better results with dropping the texture and shader quality with DOF and Motion Blur disabled...


    second run: MB OC on, XMP enabled - RAM at 3000Mhz, GPU OC Off
    2K, FullScreen, HIGH Shaders, MEDIUM Texture Quality. DoF OFF, Motion Blur OFF (VRAM usage at 1695 out of 8192)

    Results - No Crash

    FPS Avg 62.31
    GPU Temp Avg mid to high 70s, max recorded at 81
    Fan Speed/Tach (on auto profile) ~ 51% at ~ mid 1400 to low 1500 rpms
    CPU Temps Avg low 40s to mid 50s on all cores (the histogram on Afterburner showed quite a bit of back and forth between the lows and highs, a saw tooth pattern if you will, not sure if that's normal)... https://imgur.com/a/6D5pSXz


    For the third and fourth runs, I went with the default 1080p Extreme and Medium

    third run: MB OC on, XMP enabled - RAM at 3000Mhz, GPU OC Off
    1080p Extreme (VRAM usage at 3322 out of 8192)

    Results - No Crash

    FPS Avg 30.60
    GPU Temp Avg mid to high 70s, max recorded at 82
    Fan Speed/Tach (on auto profile) Approximately same as last run
    CPU Temps Avg mid to high 40s (all cores show same low 40s to mid 50s and the same sawtooth pattern, with a couple of quick spikes on all cores into the mid to high 60s at start of run).


    fourth run: MB OC on, XMP enabled - RAM at 3000Mhz, GPU OC Off
    1080p Medium
    Results - No Crash

    FPS Avg 118.92
    GPU Temp Avg mid to high 70s, max recorded at 83
    Fan Speed/Tach (on auto profile) ~ same as the other runs
    CPU Temps Avg mid 40s to high 50s (not sure if that's due to the time the pc had been running, I was givng 5 or so min between runs to record numbers fyi)



    On this last game test (MGSV again) with afterburner, I got an average GPU temp of high 60s to low 70s (74 max).

    On the PC case and fan config, it's likely much easier to show than type out a description for... https://imgur.com/8tCywf1


    The system fans are the rear exhaust, the two front intake, and the top 240 fan. I normally keep the side panel off as shown, though I do keep the front vent slots installed. HWM and BIOS show 1 sys fan though all four run. The rpms were ~ 890 rpms. I have a CM StormTrooper case, and I just maxed the fan speed on the case buttons and that's waht I saw in the BIOS. The CPU fan was roughly 1450 at idle.


    Thanks again, and hopefully that will give more useful data to look at.
  5. Well, there's a reason why PC cases have side panels and mounts for fans. Without side panel, what you have is open-air setup and having fans there makes it quite pointless. But if you'd have side panel on, the fans would create a nice airflow path inside your PC.
    In open-air test benches, GPU sits vertically where hot air can escape directly upwards. In regular PC case, hot air can't escape upwards since it would be met with GPU. And due to this orientation, GPU also relies on PC case fans and directional airflow in closed case to help cool itself.

    As far as what's going on with your PC, my suspicions got confirmed when i asked you to run 1080p benchmarks. Sadly, it looks like your GPU is dying since at current point, your GTX 1080 performs a bit better than my GTX 1060 3GB GPU. Here are my runs at 1080p Extreme and 1080p Medium:

    1080p Extreme

    1080p Medium

    Perfectly fine GTX 1080 should run 50+ FPS on 4K optimized run and not at 1080p medium run, like this one here,
    youtube: https://www.youtube.com/watch?v=tdDhxcB3UNs

    Dying GPU also explains why it can't run any of your games anymore while it ran them all just fine in the past. Though, to test it further, if you can, plug your GTX 1080 into another PC and look how it performs in there. Also, try to get your hands on 2nd GPU and look if you can get rid of your PC freezing issue while gaming.
    Without 2nd PC or 2nd GPU at hand, your GPU should be able to run your games at 1080p. You can test that out if you like, also lower in-game settings if FPS seems to be a bit low (below 60) to get enjoyable performance. Though, i can't tell for how long your GPU is able to last at 1080p until it dies completely.

    As far as what might have caused your GPU to loose most of it's performance, my suspicions lay on the high temps during gaming. Since you didn't monitor your temps, GPU probably sustained damage over period of time, up to the point where it failed to operate normally (e.g running your games). How your GPU is still going in Superposition, that i don't know but very low FPS on 4K and 1440p is clear sign that there's something very wrong with your GPU.
    Oh, when it comes to the different versions of GTX 1080 GPU, your EVGA is one of the cheapest GTX 1080 around and with cheap(er) price, there isn't much build quality in it. Which could be also one of the reasons of it's demise.
  6. Depending on it's manufacture date, Thermal issues with some EVGA cards are a known issue:

    A read at:

    http://www.guru3d.com/news-story/evga-gtx-10701080-cards-almost-all-of-them-have-overheating-issues.html

    May be enlightening

    Overclocking will worsen thermal issues of course.
  7. Mine falls squarely into the manufacture date. It looks like there is a three year warranty on the card, so hopefully I can get in touch with them after the holiday and see if we can swap it out still.

    Unfortunately, the only other card I have lying around is a GTX 670. Not sure if that would fit the bill really, but I do have another older rig I can plug the 1080 into and see if the problem persists.

    Thanks for the link cherry blossoms, edifying indeed. I'll post back on this when I've spoken with EVGA.

    Thanks again for your time, it is very much appreciated!
  8. So it was a busy past few days with the holiday and such, but I did get the GPU switched into another computer. I tested MGSV again and this time around I received zero crashes. I ran this four or five times in the last couple days and no crashes within the normal time frame I would on my main machine, or within any time frame. Things ran well for several hours at a time, each time.

    I also installed Tera and played some marathon sessions with again no problem. Obviously I was curious since I expected the machine to crash given the GPU is the suspect so far, so I ran some tests for this rig too.

    Here are the system specs for this secondary rig I used to test out the GPU:


    MSI Z77 MPower Mobo
    *Same EGA GTX 1080 SC
    Intel i5 3570K OC'd to 4.2 Ghz
    CM Evo aftermarket CPU fan
    Win 7 Pro 64 bit
    16 Gb DDR3 RAM
    128 Mb Samgsung Evo Pro 850
    Thermaltake Model SP750AH3CCB Bronze 750W PSU
    Dell 2k Gysync Monitor



    For userbenchmark, here are the results: http://www.userbenchmark.com/UserRun/9632147


    I ran Superposition again this time with no crashing, one on a 4k med settings (below) with DOF and MB on...


    FPS Avg 39.19
    GPU Temp Avg mid to high 70s, max recorded at 81


    ...then on 1080p Extreme

    FPS Avg 29.62
    GPU Temp Avg mid to high 70s, max recorded at 81


    ...then on 1080p Medium

    FPS Avg 116.97
    GPU Temp Avg mid to high 70s, max recorded at 81


    Monitoring during gaming has temps in the mid 60s to low 70s for gpu temps, GPU fans around 1100 rpms, ~ 50 for CPU temps with 100% fans, and sys fans at 65% at low 2100 rpms. Frames in each game were at 60 or more (with the expection of the MMO's popluated player areas).


    So I really expected the games to keep crashing when switching the GPU to the other rig. I know the card was manufactured during the time frame where over heating was an issue, but not sure if I should be calling EVGA on that yet or not. Maybe it's still the GPU, since the max recorded temps were still at 81 in Superpostion. I guess it's possible the seconday rig has better air flow, so the card doesn't get hot enought to crash, but is still too hot? Or maybe it's something else in my main rig that might be the culprit.

    I'm guessing without any real experieince in diagosing hardware failures and I understand Aeacus may or may not see this again, but everyone is very welcome to opine here. I really can't fathom that this problem is not fixable, I just wish I knew how to figure it out, so you have my gratitude sincerely and way ahead of time.
  9. Looking from your 2nd rig's UserBenchmark run, your GTX 1080 is underperforming. Though, 4K medium settings in Superposition provide reasonable average FPS, which is expected from GTX 1080.

    Also, your gaming temps in 2nd rig are fine. While max temp in Superposition is 81C, Superposition also fully utilizes GPU and runs it hard. It's like using Prime95 for CPU.
    To put it short: during gaming, GPU temps are fine when they stay in mid 60s, low 70s while hitting 81C at some point in Superposition. What isn't fine is when during gaming, GPU temps are mid 70s and hitting 81C since at 81C, you'll see GPU thermal throttle. Over 90C gives hardware damage to the GPU.

    What i spot from the 2x UserBenchmark runs, between your main rig and 2nd rig; in your main rig, GPU drivers are older: v391.35 while in your 2nd rig, GPU drivers are newer: v398.36. It's possible that you have GPU drivers incompatibility issue in your main rig.
    As far as GPU drivers go, i'm running v378.92 in my Skylake build. Since GPU divers doesn't give me any issue, i don't have any reason to upgrade it to the latest possible version. In other words: if it ain't broke, don't fix it.

    Though, i'm skeptical that GPU drivers is the sole reason behind the issue in your main rig. Mainly because your 2nd rig is running Win7 while your main rig is running Win10. Thing with Win10 is, is that it's updates are notorious in messing up peoples PCs.
    To test that, when making a clean install of Win10, don't update it and try if your main rig with GTX 1080 can run games as it used to be. If anything is fine, update Win10 one update package at a time and not all updates at once. After every update package, check if your PC can run games or not. While this is tedious to do; install update package, test within game, install another update package; it's only way to find out if and which Win10 update was the culprit that messed up your system.

    But before going through the trouble of finding which Win10 update it could be, if Win10 even is the culprit, i'd put the GTX 1080 back into main rig and test gaming on different resolutions first. You usually game on which reso? 4K? I'd start by setting in-game reso to 1080p and look if PC freezes up during gaming. If games run on 1080p fine, up the reso to 1440p. If 1440p also works fine, test 4K. But when even running games at 1080p causes your PC to freeze, dig out your GTX 670 and test gaming with it on your main rig on 1080p. GTX 670 should handle 1080p gaming just fine since it has a bit better performance than GTX 760, which i have in my Haswell build and where i can play on 1080p without issues.
  10. Thanks, I will take those suggestions and report back on what happened.
Ask a new question

Read More

Gaming