[SOLVED] Machine is shutting itself down after heavy gaming and need help troubleshooting

[jodriscoll]

Hoping somebody can provide a solution to this, never-ending issue with the machine I purchased...

Backstory:
I’ve been experiencing issues with “higher end” games, that after prolonged gameplay, my machine will “lock up”, at times staying locked on the frame/frames being displayed at the time and at other times restarting the machine. I purchased a smaller build (Lenovo Legion C730; spec sheet) machine, primarily to play VR in my living room, but have since been using it as a PC gaming rig. The issue most frequently appears on the titles I want to play the most (Anthem, Death Stranding, Cyberpunk 2077), almost to the point where it's ruining my overall experience with these titles. A few general observations:

• Happens more frequent than any of the games listed below while playing Cyberpunk 2077
• Locks up after long gameplay in Death Stranding
• This was a major issue with Anthem, which was the first game I truly utilized the PC for; happened sporadically, but usually after longer gameplay

Machine specs: https://pastebin.com/raw/vmuSxyS4


What actually happens:
I'll be in the middle of playing a game, then the machine will do one of the following things:

• Locks on the immediate frame that is being shown, the computers fans then go into crazy mode and spin relentlessly; this may be endless, but I usually get scared and forcefully shut down the machine (hold down power)
• Skips a couple frames, but ends up locking on a single frame, then immediately reboots
• Locks on the immediate frame that is being shown, then immediately reboots

Attempts at isolating and resolving the issue:

• FurMark: Ran for 30 minutes, GPU was running between 82c and 87c on heavy load, returns to 40-50c idle
  • Due to the temps reaching 87-88c, I replaced the (3) three stock fans with Noctua brand fans (for noise and performance)
    • Issue persists
• Memtest: Ran for 10 minutes, failed twice. First failure I had the close on failure set, second failure I captured and should be included in the album link somewhere in this post
  • Due to this error, I replaced the original memory DIMMS (SKhynix HMA82GU6CJR8N-VK 16GBx2) with the Corsair Vengeance LPX 32GB (2x16GB; CMK32GX4M2A2666C16)
    • Issue persists
• Prime95: ran for 30 minutes, CPU & MEM was maxed out, no crash or errors
• PCMark 10: ran whatever came with the free version, nothing crashed or locked up due to errors
• 3DMark: ran the D12 Time Spy benchmark and nothing crashed or failed
• AIDA64 Benchmarks: ran all benchmark tests and all passed without any crashes
• MSI Afterburner Overclock: test reported a 92% confidence score
• Valley Benchmark 1.0: ran for 30 minutes, no issues or crashes, but the experience put me to sleep
• Heaven Benchmark 4.0: ran for 30 minutes, no issues or crashes, but the experience was relaxing
• Onboard Lenovo Diagnostics: when booting, pushing F10 will boot us this alternative UI, which takes about 9 hours to complete (no joke), but after completing, everything passed and nothing failed the tests
• EA Support: I had my issues with Anthem escalated on a phone support tech, who is turn had me dig through some system logs (Event 6008) and, in short, said "it wasn't the game and you needed to have your machine looked at"; helpful, I know...
  • I entered the Windows Event Viewer, located the events where the system shutdown, he viewed them, tried to isolate the problem, then dismissed it as it was not the games fault
• Reinstalled Windows OS: on the original on-board SSD, as well as installed on a new SATA SSD
• Drivers: Updated all I’m able to locate updates for relative to hardware
• Drives: Samsung magician says all drives are operating at expected speeds

Screenshots:

https://imgur.com/a/18L4OC9

Endgame:
I'm looking for any possible techniques or solutions for resolving this problem. I've considered building a new machine, with better ventilation, which I'm sure I will end up having too, but this machine is in great condition and to "write it off" for an issue that only happens under the most "demanding games", is, unfair? I've tried to do a ton of research, reading, digging through previous threads for advice/solutions, etc., and I'm currently at the current mind state:

• The stock PSU is not sufficient for the upgraded components, under heavy load, and needs to be upgraded. The "stock PSU" is quite literally built to fit the case, so through my research, I'm considering the following: Fractal Design Ion+ 760P 80 PLUS Platinum Certified
  • With the fan almost LITERALLY sitting up against the side panel, I was a bit concerned, but, due to the PSU needing a certain form factor (due to the case), I'm limited
• The airflow going into the front of the machine is not producing enough "cool air", that in turn is digested by the GPU (from the top)
  • Purchased a small USB fan to utilize during gaming sessions to push more air through the front; will utilize it for a "nausea avoider" when hooked up to the VR
• Give up on this machine being able to handle the higher-end games I want to play and I'm starting to save for a (non mini-itx) build; keep this for the VR/home media server

Any help would be greatly appreciated. I'll try my best to respond as promptly as possible.

 

[ironstem]

From the looks of it, you've definitely overheated everything while doing heavy gaming. I think the pc is shutting itself off because its detecting problems. If you haven't had good airflow within the pc, then that's already bad enough. Obviously, that causes overheating. But the PSU not being sufficient enough? That causes stuttering and potentially frying components within the system. You've definitely screwed something up man.

 

[jodriscoll]

From the looks of it, you've definitely overheated everything while doing heavy gaming. I think the pc is shutting itself off because its detecting problems. If you haven't had good airflow within the pc, then that's already bad enough. Obviously, that causes overheating. But the PSU not being sufficient enough? That causes stuttering and potentially frying components within the system. You've definitely screwed something up man.

Is there a sure-shot way of verifying that the temperature/overheating is the exact cause of the system forcefully shutting itself down? As a general note, this is a prebuilt machine by Lenovo (C730), which I did not adjust anything from the factory until recently. Meaning, this issue existed within a month of purchasing the machine. I guess what I'm trying to say is, I haven't modified anything until recently (within the past 2 months), in which I actually (assumingly) upgraded the 3 fans within the case; example of a similar upgrade.

The PSU mention was based on research from similar posts where gamers machines were shutting down immediately if the peak power exceeded the total PSU capacity. One user (I'll dig up the link on Tomshardware) was able to isolate it to the PSU failing to generate enough peak power; the Linus youtube channel briefly covered this happening during a video as well (link, happens @ 07:35 ).

To be clear, I don't believe I've "screwed" anything up, but in fact, have only improved it (upgraded fans, memory, reorganized the entire inside of the case to expose more opportunity for airflow, etc.). If somebody could provide a path of solidifying the overheating is the exact issue, I'm happy to conclude this annoying issue and simply, buy a different ITX case with better airflow.

I'm fully aware that "stress tests" software does not simulate "real-life conditions" as actually playing a game for hours, which makes sense that the machine would nearly pass every tests/benchmark/stress process I've put it through. I'm just not aware of any other methods of elimination and isolation of this particular type of machine problem outside of an observation of "the temperature of the GPU right now looks like it's hot". I mention this mainly due to the machine experiencing the issue when the GPU was at 82c, 85c and 88c, which makes it hard to think that it is reaching a temperature and immediately shutting down.

Thoughts?

 

[DSzymborski]

I'd absolutely want to rule out the PSU here. A 500W unidentified PSU isn't exactly an auspicious sign. You've done a great job documenting what's going on and what you're trying to do, but we still don't know anything about the PSU other than the wattage and this is one of the most important bits of info given your symptoms.

 

[jodriscoll]

I'd absolutely want to rule out the PSU here. A 500W unidentified PSU isn't exactly an auspicious sign. You've done a great job documenting what's going on and what you're trying to do, but we still don't know anything about the PSU other than the wattage and this is one of the most important bits of info given your symptoms.

Much appreciate your response. I delayed a bit to track down some information and produce a short video (link) showing a rather "interesting behavior" with Cyberpunk 2077 that continues to push me towards thinking this problem is relative to the PSU.

To briefly tl;dr the video - after playing the game (Cyberpunk 2077) for about an hour, with the case partially opened to maintain an average 82-85c GPU temp, I noticed that when opening the game inventory menu, then navigating around items contained within, the FPS was dropping and there was an odd "buzzing" sounds coming from the machine each time the FPS was dipping. Could be game specific, but it was audibly distracting with the machine opened, which was concerning.

Anyhow, the PSU is indeed the most "mysterious" part of the stock build from Lenovo. I also haven't been able to track down an easy way of verifying the peripheral without actually removing it. Unfortunately, as you can see by this image (link), it doesn't immediately present the "label" to the user when opening the panel housing the PSU - here is a picture of the label. I did locate a conversation on Tomshardware (link) where a few users discussed the model(s) and the aftermarket units they replaced the stock unit with.

Random inquiry:
Is there any recommended tools/applications/software on finding out a PSU's "technical tidbits" without physically removing it? I only ask as this case was a PITA to get the PSU out, then a PITA to put it back in and have everything "fit nicely", heh.


UPDATE:

• Older video of what the machine was doing when it locked up (this is within months of purchasing and letting it idle in the game Anthem)
  • This is an example of the machine "Locks on the immediate frame that is being shown, then immediately reboots" (mentioned in my initial post)
• Another older video of what the machine was doing while "stuck in a locked up" situation (again, same general time as above, within months of purchasing)
  • This is an example of the machine "Locks on the immediate frame that is being shown, the computers fans then go into crazy mode and spin relentlessly"
  • Note, when I say "can you hear it" in this video, I'm referring to the sound after my statement, not my puppy yawning at the very moment I said that

 

[jodriscoll]

@DSzymborski - PSU information listed below (screenshot in the reply above), hope that helps!

• LITEON PS-8501-2
• 500W
• 100-240V
• 50-60hz

 

[DSzymborski]

@DSzymborski - PSU information listed below (screenshot in the reply above), hope that helps!

• LITEON PS-8501-2
• 500W
• 100-240V
• 50-60hz

Ouch, a junky old group-regulated PSU is not something one likes to see in a build with an RTX 2070 (it's not even a real 500W PSU, it can barely pass 400W on the +12V rail, which is atrocious by post-1999 standards). The PSU included was selected to do one thing: survive one year at the original specs so that Lenovo doesn't have to pay to fix anything.

As much as a pain the form factor is, this PSU should have been replaced at the same time as the 2070 , so getting it out of the mix when there actually are problems is a very obvious next step since it needs to be done anyway.

 

[jodriscoll]

Ouch, a junky old group-regulated PSU is not something one likes to see in a build with an RTX 2070 (it's not even a real 500W PSU, it can barely pass 400W on the +12V rail, which is atrocious by post-1999 standards). The PSU included was selected to do one thing: survive one year at the original specs so that Lenovo doesn't have to pay to fix anything.

As much as a pain the form factor is, this PSU should have been replaced at the same time as the 2070 , so getting it out of the mix when there actually are problems is a very obvious next step since it needs to be done anyway.

Thanks for responding! I figured there may be issues with this PSU, but to clarify, I haven't replaced the GPU that the machine (when purchased) was shipped with (saw something about replacing). I don't mind upgrading the PSU, I'm actually on the Newegg.com waiting list for the Fractal Design Ion SFX 650G 80 PLUS Gold, but I could purchase the Seasonic FOCUS SGX-650, 650W 80+ Gold immediately.

Thoughts?

Side conversation:
Also, if you were able to view the video above (where there is an odd buzzing sound while in the game menus), do you have any idea what that could be? I just stumbled on the Intel XTU software and during the benchmark tests of the CPU/Power, I heard (faintly) a similar sound.

UPDATE:
- Added a bunch of Argus Monitor screen grabs, as well as Intel XTU screen grabs (hope that helps, help me!) to the album link above (link, jic)