PC Crashing Harder than the Titanic

[littleCERBERUS]

Alright ladies and gentlemen. I'm stumped. My computer keeps crashing (instant turn off, as if I had pulled the plug) and starting itself back up again two seconds later. However, it's mostly random! I've only noticed 3 programs that crash it consistently, but of those three, only one is a replicate-able crash.

 

Photoshop CC 2018 (newest version) crashes it, but only once in a while. Yesterday I used it for hours with no issues, but it's crashed 10 seconds in before.

Path of Exile (also up to date) crashes it, but VERY randomly. Sometimes I can play for 4 hours, sometimes I can play for 4 minutes. No telling.

Assassin's Creed: Odyssey crashes it, and this is the only consistent one. The opening cutscene of this game crashes my computer within about 15 seconds or so. I've never made it more than 3 lines of dialogue into that cutscene (6+ attempts with different settings and such).

 

Other games/programs have crashed it before too, but it's totally random. For instance, I've played probably 20 hours of Assassin's Creed: Origins in the past 3 days (yesterday being exceptionally long), and never had that game crash my computer. I don't think it's a temperature issue, because, like I said, I played Origins for probably 8 hour straight yesterday, and never had a single problem.

 

OS - Windows 10 Pro version 10.0.17134

Motherboard - Gigabyte Z97X-SOC

Processor - Intel Core i7-4790K

CPU Cooler - Corsair H100i (original, not v2)

RAM - 16 GB (2x 8GB sticks) G. Skill Ares DDR3 (don't remember and can't seem to figure out the MHz, sorry)

GPU - EVGA GeForce GTX 1080TI FTW3 (until 2 months ago, I had a Gigabyte GTX980, and was still having these issues back then)

PSU - EVGA 1000 G 80+ Gold

Hard Drives - (not sure of most of the details unless you can tell me where to find them) 1x 120GB (or 128GB?) Samsung SSD, 1x 1TB 3.5 drive, 1x 2TB 3.5 drive, 1x 3TB 3.5 drive (which I admittedly removed from its enclosure and hooked up internally, because it used to be a USB external drive).

Case - Corsair Graphite 760T Full Tower

SMBIOS Version 2.7 (is this that BIOS version thing that was asked for? I know NOTHING about BIOS)

 

I also have NOT overclocked anything, and all of my drivers are up to date (I believe), but this has happened across MULTIPLE MULTIPLE driver versions, so I'm pretty sure that's not the issue. This has been happening for quite some time (6+ months) now.

 

Please Help.

[knightslugger]

over current protection?

[littleCERBERUS]

1 minute ago, knightslugger said:

over current protection?

Please excuse my noobishness, but I'm not sure what you mean. If you explain, I'd be happy to answer!

[Dedayog]

Yep, over current protection.  Had that issue with my EVGA 650 P2, due to GPU draw spikes.  RMA the PSU.

[RollTime]

1 minute ago, knightslugger said:

over current protection?

He has a 1000 watt PSU, I doubt he's exceeding it.

 

Just now, littleCERBERUS said:

Please excuse my noobishness, but I'm not sure what you mean. If you explain, I'd be happy to answer!

Your PC may be pulling more power than your PSU can handle. Have you tried another PSU? Have you tried another OS?

[littleCERBERUS]

OH. it wasn't a question, it was a PROBLEM. So my PSU is bad, huh? Well that stinks.

[RollTime]

1 minute ago, jstudrawa said:

Yep, over current protection.  Had that issue with my EVGA 650 P2, due to GPU draw spikes.  RMA the PSU.

With a 1000 watt PSU? 

[littleCERBERUS]

Just now, RollTime said:

He has a 1000 watt PSU, I doubt he's exceeding it.

 

Your PC may be pulling more power than your PSU can handle. Have you tried another PSU? Have you tried another OS?

I haven't tried another PSU (don't have a spare), and haven't tried another OS (same reason).

[RollTime]

Just now, littleCERBERUS said:

I haven't tried another PSU (don't have a spare), and haven't tried another OS (same reason).

Try Linux. And maybe you know someone who can lend you one?

[littleCERBERUS]

8 minutes ago, RollTime said:

Try Linux. And maybe you know someone who can lend you one?

I really don't. And even if I did, I wouldn't know how to install/use it, and then wipe it and get Windows back on it (because I really need Windows).

[Fasauceome]

Have you done a memory test? The random nature makes me feel like memtest could help you verify what's wrong

[littleCERBERUS]

3 minutes ago, fasauceome said:

Have you done a memory test? The random nature makes me feel like memtest could help you verify what's wrong

I have not. How would I go about doing this?

[Fasauceome]

8 minutes ago, littleCERBERUS said:

I have not. How would I go about doing this?

Memtest can be downloaded online, and so can Aida64 which has a memory stress test.

[littleCERBERUS]

38 minutes ago, fasauceome said:

Have you done a memory test? The random nature makes me feel like memtest could help you verify what's wrong

Ok. Downloaded Aida64. Ran the System stability test (default has Stress CPU, FPU, cache, and system memory all selected). Literally the MOMENT I clicked "Start", the system died. Once it booted back up, I ran Aida64 System Stability Test AGAIN, but this time ONLY checked "Stress system memory". It ran fine for 15 minutes. No issues. I'm about to "one-at-a-time" the CPU, FPU, and cache (in that order) to see if I can replicate the insta-crash. If I can replicate the insta-crash, I think I'll have a better idea what's wrong. About to try CPU. brb.

[Dedayog]

1 hour ago, RollTime said:

With a 1000 watt PSU? 

It's not the wattage amount, it's making sure the draw is in all the specs for the protection.  Could be a tiny voltage issue, who knows.

 

Talking with EVGA, the most likely culprit is the GPU since it's draws can fluctuate so much.  But anything can really be an issue, just depends if it's out of parameter ever..

 

Another issue with the EVGA units I looked up was that Eco mode wasn't able to come up fast enough at times resulting in heat buildup that damage units over time.  There's a few forum posts about it but nothing concrete.

 

http://www.tomshardware.com/answers/id-3025162/evga-850w-eco-mode-causing-crash.html#r19739619

[littleCERBERUS]

Ok. Just ran the CPU stress test for 15 minutes, again, no crash. Guess I'll try FPU now.

[littleCERBERUS]

8 minutes ago, jstudrawa said:

It's not the wattage amount, it's making sure the draw is in all the specs for the protection.  Could be a tiny voltage issue, who knows.

 

Talking with EVGA, the most likely culprit is the GPU since it's draws can fluctuate so much.  But anything can really be an issue, just depends if it's out of parameter ever..

 

Another issue with the EVGA units I looked up was that Eco mode wasn't able to come up fast enough at times resulting in heat buildup that damage units over time.  There's a few forum posts about it but nothing concrete.

 

http://www.tomshardware.com/answers/id-3025162/evga-850w-eco-mode-causing-crash.html#r19739619

So wait....is it the PSU or the GPU? It sounds like you are implicating the PSU, but then you said that EVGA blames the GPU? Just looking for clarification.

[Dedayog]

2 minutes ago, littleCERBERUS said:

So wait....is it the PSU or the GPU? It sounds like you are implicating the PSU, but then you said that EVGA blames the GPU? Just looking for clarification.

PSU trips protection, but was infact damaged.  Per EVGA.  

 

They said it was the GPU doing it.  In my experience with EVGA, I no longer take what they say as gospel or even truth.  

 

I got another new RETAIL psu from EVGA (retail is important, since it means they didn't send some piece of shit no-fixed refurb) and I've been fine for 6 months now.  

 

Edit:  If you want, to be safe, RMA your 1080Ti as well.  They'll test it.  Hopefully.  It's EVGA too   For my point, here are the NUMEROUS psus they sent me that did not work, they even admitted that 2 of them didn't work when they tested them after I sent them back, pissed.

[littleCERBERUS]

OK! RESULTS!

 

Aida64 "Stress FPU" test resulted in ANOTHER INSTANT CRASH. So this seems to be the culprit. So my next question:

 

What's "FPU" and how do I fix it? lol.

[Fasauceome]

14 minutes ago, littleCERBERUS said:

OK! RESULTS!

 

Aida64 "Stress FPU" test resulted in ANOTHER INSTANT CRASH. So this seems to be the culprit. So my next question:

 

What's "FPU" and how do I fix it? lol.

FPU is floating point unit, it's part of your CPU. it may indicate that your CPU is damaged if the FPU can't handle a stress test.

[littleCERBERUS]

11 minutes ago, fasauceome said:

FPU is floating point unit, it's part of your CPU. it may indicate that your CPU is damaged if the FPU can't handle a stress test.

Even though the CPU stress test ran fine?

 

Also (just my need to understand things), if this...FPU thing....is the problem, why would that trigger only seemingly randomly during these other programs (and fairly consistently during AC: Odyssey)?

 

Also, is there any way to be SURE this is the issue? buying a new CPU isn't exactly a cheap fix.

[Fasauceome]

1 minute ago, littleCERBERUS said:

Even though the CPU stress test ran fine?

 

Also (just my need to understand things), if this...FPU thing....is the problem, why would that trigger only seemingly randomly during these other programs (and fairly consistently during AC: Odyssey)?

 

Also, is there any way to be SURE this is the issue? buying a new CPU isn't exactly a cheap fix.

You can continue to do extensive testing with software, or you can try your hand with different hardware like a new ram kit or motherboard, lot cheaper than a motherboard.