In August I bought this https://www.microcenter.com/product/638791/powerspec-g468-gaming-pc PC. I also paid an additional few hundred dollars to upgrade my ram to 64GB. The tech said that some of my ram sticks were malfunctioning so he swapped them. I didn't think much of it.
As of January 6th my PC started bluescreening within 10 minutes of playing a game. It can be any game, Cyberpunk, PUBG, CSGO. I looked at every possible solution and updated every driver I could think of. I ran Memory Diagnostic (no shown issues). I finally resorted to resetting the PC in case it's a driver issue and I'm still having the same problem. This is a new computer that I paid $3K for, I never in a million years expected it to crap out on me right at the 6 month mark....
Here are some of the error logs pre-reset:
Same Blue Screen after Reset:
For the 7th those are not the full errors. At one point after reboot it asked me if I want to run a repair, which I did. That seemed to remove some of the software I had installed and clean out some of the dumps. Regardless it's always the same error.
Here is info using WinDbg:
Looks like memory instability. The original kit from the factory was 3600 C16 GSkill kit. I'm showing in your purchase history a 3600 C18 kit. We need to figure out what this RAM is trying to run at, so far as the timings. Please run CPU-Z and screenshot the memory and SPD tabs. On the SPD gab, please capture either slot 1 and 2, or 3 and 4 so we can see both modules.
Below are all the slots in case one them varies (although I think they're all the same). It's strange since it worked fine for so long but now almost to the clock 8-10 mins after launching a game the computer crashes. Also, thanks for the help!
Looking at that, it looks like all four modules were replaced. That's good, preferably we don't want to mix IC's. Please also screenshot the "Memory" tab. This tab shows up the profiles, but the Memory tab will show us what you're actually running at.
Here you go:
Uncore Frequency is randomly jumping around from 799.9 Mhz to 4299.0Mhz and back.
You're running at 2666 CL19 instead of 3200 CL18. Probably 1.2V. Lets do this.
Reboot and tap "Del" to enter the BIOS. Go to AI Tweaker - AI Tuner = XMP II. Save and exit. Check CPU-Z again and confirm the settings. Should show 1600 CL18. If so run your windows memory diagnostics.
All I had was AI Overlock Tuner. I didn't just overlock my PC, did I? Also I'm not sure where you are seeing 2666 CL19/3200 CL18 so I attached the CPU-Z.
The CL did change to 18, but I don't see 1600 anywhere.
Here is what I got after saving with XMP II:
Sorry, that was a typo. I meant 1800. This looks good. You loaded the XMP profile, you're overclocking the memory above a JEDEC standard but this is what it's rated to run at. We ship it with the XMP profile loaded. Loading the XMP profile would not impact your warranty. Lets run the windows memory diagnostics and see if it detects any errors.
Woah. That took over two hours to complete. The results show no error. A note about the BIOS settings, I think I might have had these #s or similar ones from the start. Once my bluescreen crashes started to happen I tested everything, including resetting BIOS to default. I'll launch a game and see how things go. 🤞
Thanks a ton for helping me through this Mike!
Unfortunately the issue still persists. The interesting thing now is that the crash is caused by another file according to debug: e2f68.sys.
The big difference is after the computer crashes and restarts I can login immediately without it attempting to boot and then giving me a Blue Screen or forcing me into Recovery Mode. The bizarre thing is these crashes ONLY happen when playing video games. Zero issues when browsing the web. Could it be video card related?
I updated my Intel Drivers, launched the game again, played roughly the same amount and then crashed to desktop. Then I open BSOD by instinct and the computer BSOD's. Something is really really wrong with this computer. After restart the debug error went back to what it was (ntoskrnl.exe).
e268.sys references the network controller, but I don't believe it's significant. You've seen various errors and it's just random, which I'd usually attribute to memory first, but we've ruled that out. Now the freezes make me think GPU, and that's what's under load when you 're gaming. But you're not logging display driver errors and it generally wouldn't be attributed to random BSOD's, the BSOD should point towards an Nvidia driver. If it was the NVME drive, typically the drive becomes unresponsive. You wouldn't have any minidumps to look at. The common link that involves both items is the PCIe bus. Lets try this.
Enter the BIOS. Advanced - Onboard Device Configuration - PCIEX16_1 Mode = GEN3.
Save, exit and test a game again.
Adding this as it's getting late here. The above suggestion with changing the PCIE GEN is a longshot, not likely to offer a solution. It may change the behavior and give us more insight into the root cause, but all roads with what you're seeing point towards a hardware issue. Unfortunately either way we're going to need to take care of this for you under the 1 year manufacturer warranty.
Assuming the above change has no effect at all, the other component that explains your symptoms and what I would think is most likely is the power supply. I'd assume it's failing under load. Voltage drops and creates your instability, but not enough to trigger a shutdown.
Got it, I'll stop tagging you. A few notes about this experiment.
This is really unfortunate because 3 weeks ago I moved to NC and closest MC is 4.5 hrs from me. 😔
I'll try and watch HWIMonitor and see if PSU drops before crashes.
" Unfortunately either way we're going to need to take care of this for you under the 1 year manufacturer warranty."
I know you meant fortunately. I'd be so mad if I paid 3 grand for a chrome book. 😜
nvm. My posts keep vanishing.
You can tag me, I'll certainly get back to you when I can. This might tell us a little bit. Was the BSOD happened 5 seconds after logging in recorded? Do you have a minidump from that we can see in windbg?
The black screen flickering sounds like a display issue, but if it was we should be seeing Nvidia display driver crashes. Which points me back towards another factor causing instability. Power supply just makes the most sense. Here's something we could try. Run Prime95 for 20 minutes, small FFT's. If it doesn't crash, try furmark for 20 minutes. 4k, x8 AA. If that doesn't do it, run them at the same time.
Yes, I meant unfortunately from the standpoint that it would need to be repaired and would result in downtime.
Sorry about that. I saw your posts but they were being marked as spam automatically. Marked you as verified and that shouldn't happen again.
One other thing that might be telling here. The next time it crashes, since it sounds like it's pretty consistently hard to get it to boot back up right away. Requires at least one restart. Turn the system off, pull the power cable. Let the system discharge for 30-60 seconds then try it again. See if that averts some of the post crash symptoms.
Don't see a capture in the minidump. Below I have a few phone pics I captured during the whole process. I'm contemplating buying a PSU because driving to a MC store, dropping off the PC, & coming back a week later to pick it up would be like 16 hours of driving. That's insane. I might as well buy new hardware. lol
Tried to get in contact with a store today but never heard back. Ideally I should be able to ship an item to the store while it's under warranty.
Thanks for the verified status! I was starting to think my internet is also broken. 😅
Wouldn't be a bad idea ruling out any external factors that would effect the power supply. Power cord if you have a spare. Testing another outlet in the house if possible. It's definitely difficult to pin this down with what you're seeing. It's causing this warm boot issue after the crash, you're seeing the board fall back into fail safes. The BSOD above is interesting. The drive is accessible, but it can't read a file, but does boot the second time. 90% of the time, that indicates RAM. Clearly it's not. Power still makes sense, but it could be something else. Board VRM as an example. Stressing the CPU/GPU individually may tells more.
Yup, one of the first things I did was unplug all the devices and test out each individually. I also changed the power cable and the outlet. I'm going to stop replying now so you can get some rest. Thanks again for all the help today Mike!
Sounds like you've covered all your bases and we're going to need to swap some parts out to confirm. Looking at your last few messages it does seem once this system starts crashing, shutting it down, discharging it seems to return it normal until you put it under a significant load again(Game). So you have to think power or heat. I saw you're 28C in the BIOS and given the excellent information you've provided and that you're using HWINFO64 I'm sure that's all been looked at thoroughly. And with temps you'd likely see a performance dip and a throttle prior to the crash. That leaves power, so the power supply is the easiest place to start, but there are other components that could be the problem. As always with troubleshooting, you work your way back to the board.
What build is your G468?
When you say build what do you mean? The specs? I've attached system info below as well as a video I recorded of HWINfo Voltages in real-time
as the crash happens. Previously I thought it was usually after 8-10 minutes but now it looks like it's almost always 5 minutes.
Video (you can skip to the end to see final voltage/crash) : https://www.youtube.com/watch?v=GvG2kNGTixo
Edit: I also tested heat and there weren't any significant changes during the game running/crash.
The full build should be the same as this link apart from the ram:
Build number should be on the rear chassis, on the serial number sticker. It'll allow us to reference the component changes on that particular model. Not likely to relevant, but just in case.
Diagnostically our next step here is to swap out the PSU and stop there. Have you heard back from the store?
Nope. I called warranty # and spoke with a representative. He tried calling them and transferring me but it didn't work so he said that he'll e-mail them and I should expect a call by 3pm. Never received a call. I'll e-mail them directly now. Below is the sticker of the build.
One additional note, it seems the crashing only happens when running high end games. For example if I play something like a visual novel I don't crash.
Just changed the GPU on that build. Still comes down to sounding like a power supply issue, it's failing under a moderate to heavy load. The only thing I can suggest without replacing a component is what I suggested before. Try Prime95 with small FFT's, to see if you can replicate it with CPU load only. Then try Furmark which will put a heavy load on the GPU to see if one or the other triggers a response. If Prime95 caused it, but Furmark didn't, I'd lean more towards it being a motherboard VRM issue. If Furmark causes it, but Prime95 doesn't. Or it happens when you run both at the same time only, I'd lean towards the power supply.
I'll check on your ticket Monday morning either way and do what I can to make sure things are moving along swiftly.
I'm not familiar with the software you mentioned so wanted to make sure I'm downloading the right stuff.
You found the correct pages. For Prime95, select Small FFT's, that should give us a pretty heavy load on the CPU only. For Furmark, select 4K and max AA at 8X. That'll push your card to 100% and we'll see what happens.
Okay I ran both tests and it crashed during both. Below is the sumary
Prime95 - I left it running for about an hour. When I came back my computer was turned off (meaning it crashed at some point). I arrived at 6:56 so it looks like it crashed 1min before I checked. The minidump can be seen below:
Furmark - For this one I physically saw it happen. When it was ~50% into the process I blue screened on my right monitor and my left monitor become the same sharp high contrast/pixely color I witness whenever I get a crash while playing a game (you can see it on the picture on the left).
It then rebooted and gave me a system recovery option. I restarted with the power button and BIOS Safemod popped up. I told it to boot in safemode but it reverted back to the the cycle of my monitor power button turning on and off and the motherboard doing power cycles while the screen remains black. I had to unplug the power cable multiple times and keep trying until I was eventually able to start.
On Prime95, the system was running with a black screen or completely off? Off indicates it encountered an emergency shutdown condition. That's either power or heat. We've ruled out heat.
I wasn't aware of the pixelated screen which says video card. Now again, this could be back to power, which is supported by the Prime95 results above. Lower odds chance is a video card issue. Which only makes sense if it's effecting other PCIe devices, namely your NVME drive.
We need to get it in for service. If you want to try to fix this on your own, I'd start with the power supply. If you go this route, keep the old power supply in case there are any additional warranty issues down the road.
The system was shut down. As in it was offline and I had to press the power button to boot it up.
Okay, I'm chatting with support now. I need to figure out a way to get my PC to VA without having to do 16hrs of driving.
Alrighty support sent a request to be contacted so now there are two pending requests. Hopefully I hear back by Tuesday. Once again, I really appreciate all the help you've given here! Learned a lot.
Alright. We apologize for the inconvenience and the downtime with your new system. I'll keep an eye on this case, we'll make sure it gets taken care of.
I just heard back. It looks like the shipping might get covered after all. I feel bad for jumping the gun, but as I mentioned, this has been an incredibly frustrating situation. Hopefully all goes well and I can go back to enjoying my PC!
A little update on this. I mentioned earlier, but the closest Microcenter to me is in Fairfax (5 hrs away). Since the likely culprit is the video card, I've been keeping in touch with the Fairfax store to see if they have any 3080s in stock. Three weeks later and they haven't been stocked (or so I'm told). I asked the store if I can ship them my computer. The tech confirmed that I can but he wants me to pay for the shipping to and from the store. So we're talking at the very least $200+.
I paid $3100+ for this faulty machine and in 6 months it blue screened on me. Now to repair it, Microcenter wants me throw in more money. I expected them to at the very least cover the shipping back, but nope. I really appreciate your help @TSMikeW, but this whole experience has been a nightmare. For all I know the issue is something minor like a faulty ram stick and I'm paying crazy shipping fees for a fraction of the actual repair cost. Not the mention that the actual machine was sold to me with faulty ram and the only reason I found out about it is because I paid to have extra ram installed while buying the PC. What if they ship it back and break something else? Am I also paying for that?
I asked the tech if I can get a diagnostic check locally and get the parts shipped to me but he hasn't responded yet. He did dodge my shipping question the first time I asked so possibly same scenario.
The To Obtain Mail-In Service Under this Limited Warranty section in the warranty is a complete scam. They just tell you to contact the store & the store tells you to pay out of your pocket. It's there to make you think it's part of your warranty but as I've mentioned above it's not. 👎️
Submit photos and a description of your PC to our build showcase
See other custom PC builds and get some ideas for what can be done
Services starting at $149.99