Is it really happening?
Is it really happening?
I have been trying to figure out why my PC is crashing when I play games. It seems like this happens in about 30 minutes for any game I play. Before that, I was trying to fix a Kernel 41 problem and decided to reset my computer because I thought a program caused the crash. I did a full reset, deleted all my files, and reinstalled Windows. But nothing worked and the first game I installed after the reset also crashed hard like before. I am starting to think it is a hardware issue now, but I don't know where to start. Could someone please help me with the event viewer log from the crash and the minidump file? I do not know how to read that data. Here is the link: https://drive.google.com/file/d/1TfkAe_q...drive_link Any help would be very much appreciated. Thank you.
Hey, I need to know what the whole machine's specifications are, like which power supply unit it uses and its specific part number. Could you also tell me exactly when this power supply was made, how long it has been running, and if it came brand new or if it was used, refurbished, or bought second-hand?
Hello, yes sorry I should have included that in the first post. Windows 11 Mobo - Gigabyte x590 Aorus Xtreme CPU - AMD 5950x AIO - Aorus Waterforce 360 rad GPU - 3090 FE RAM - 32 GB DDR4 4266 Trident Z Storage - 1 tb m.2 PSU - Seasonic 1300 Platinum This was a new build that I bought off of FB marketplace from someone who loves tech very much. The listing said all parts were brand new, so I think the power supply was definitely new too. Also, I tested the RAM twice with memtest86 and found no problems either time. I am really leaning towards thinking it's a RAM issue right now though.
That RAM speed is not really possible in a perfect match. Try setting it to something slower like 4000 MT/s or just reset them back to JEDEC speeds so you can test properly. Did you do a clean install of Windows? What is the exact model of that Seasonic platinum 1300W power supply? Have all the CPU power cables in the top left of the motherboard installed? Do you have three 8-pin cables going to and from the PSU plugged into the 3090 FE?
Power delivery. Either overheating CPU/GPU or PSU issue. Rarely, MoBo VRMs or main electricity grid. So, what are under load temps for CPU and GPU? You can use HWinfo64 in sensors mode for seeing temps: https://www.hwinfo.com/download/ There is only one: Seasonic PRIME PX-1300, specs: https://seasonic.com/prime-px/ RTX 3090 FE doesn't use 3x 8-pin PCI-E power cables. Instead, it uses 12-pin PCI-E (forerunner of 12VHPWR connector), specs: https://www.techpowerup.com/gpu-specs/ge...3090.c3622 So, OP must use an adapter. PRIME PX-1300 was released in 2019. So, it would be safe to assume that your PSU is at least 6 years old. Now, PRIME PX-1300, with 12 year warranty, is a great quality PSU. (Oh, PSU's warranty did transfer over you as well, since Seasonic is only one who offers transferable warranty. But this is a secret 🤫 source , and with this, when need be, you can RMA the PSU.) In most situations, great quality PSU should not act up. Unless previous owner abused it somehow. Still, PRIME is robust and i'm doubtful that PSU would be an issue. Nevertheless, PSU issue can not be ruled out since there are lemons even among the best. And kernel ID41 is power delivery error, pointing towards PSU to be one possible source of the issue. So, to rule out PSU's issue, do try with 2nd, good/great quality, known to work PSU. If all is well with 2nd PSU, you can RMA your PX-1300 and get working replacement, whereby after that, you're golden. Btw, welcome to the PRIME family! (I'm running two PRIME units myself, but mine are TX-650 ones, powering my Skylake and Haswell builds, full specs with pics in my sig.)
I know what you're saying about needing three separate 8-pin power plugs for a PCIe adapter. Just making sure? Are those cables meant to be daisy-chained together from one end all the way to the other, or is there a specific setup needed like connecting them directly to the PSU and the graphics card instead? I've seen many different versions of these PSUs over the years.
The memory frequency is now set back to default, so my BIOS was reset during troubleshooting. All the DRAM settings are set to auto, and the RAM runs at 2133 MHz just like the BIOS default says. I only copied those specs from the owner while writing this out, sorry for any confusion there. A clean install was already done. The PSU in question is the PX-1300. All cables are properly connected. I've been using HWInfo to watch temperatures since I first thought that might be what caused it. Since then, I have been running the data log while playing games and checking it after every crash. I haven't seen the CPU or GPU get high temps or spike up in the logs. In a game before a crash, both showed never hitting 70 degrees Celsius. Thanks for helping with the warranty! I was thinking the PSU could be to blame but didn't know how to test it myself. Maybe someone else will swap it out for a while so we can check that. I'll try to make that happen!
Well, you didn't say that you're aware. What you said, was direct connection from PSU to the GPU, without any kind of adapter (or power cable extensions in that matter). So, i corrected that. GPU 12-pin PCI-E adapter does use 3x 8-pin PCI-E cables, but PX-1300 has only one PCI-E power cable that has two 8-pin connectors per 1 cable. All other PCI-E cables (6 of them) are with individual connector. So, even when OP used that one daisy-chained power cable, i doubt it being the culprit. Yes, there are several PRIME PSUs made over the years, but only one 1300W 80+ Platinum unit. Original PRIME series, Gold and Platinum ones, went as high as 1200W. Titanium one, as high as 1000W. I have the original PRIME, SSR-650TD. Successor of PRIME, PRIME Ultra series, any efficiency, went as high as 1000W. I have one PRIME Ultra as well, SSR-650TR. 3rd iteration of PRIME series, GX/PX/TX (current lineup), has Gold (GX) going as high as 1000W. Platinum (PX) and Titanium (TX) go as high as 1600W, while also including 1300W models. So, there is no way that OP can have any other PRIME 1300W Platinum unit, other than one and only PRIME PX-1300. Only way to test a PSU, is with another PSU (replacement). Now, there are PSU testers out there and there is also paperclip method, but they all are useless. All those testers (including paperclip method) does, is turning on the PSU and that's it. Testers can also show voltages when PSU is turned on, but that with 0 load on the PSU itself. It's like you turning on the car engine and when engine starts, you instantly assume that the car drives just fine, without ever doing the test drive (feel how car handles, test brakes, hear if engine sounds normal etc.). So, no. Just turning on the PSU, without putting any load to it, doesn't mean PSU works fine. Proper PSU testing equipment costs thousands, if not tens of thousands of dollars and is very expensive. E.g here's TH's PSU testing equipment and methodology: