PC building woes
I recently bought a new processor, motherboard, RAM and CPU cooler to upgrade my computer. Here’s the list of specific parts:
Intel Core i9-12900K 5.2 GHz
Asus ROG STRIX Z690-F GAMING WIFI
G.Skill Trident Z5 RGB DDR5 6000MHz 32GB 2x16GB CL40
Noctua NH-D15 chromax.black
The RAM sticks pass the built-in MemTest64, the processor withstands the Cinebench test, but when I tried to use my VR headset, the computer shut off. I downloaded 3DMark, suspecting there was a problem with the GPU (a GeForce RTX 2080 Ti that has been working fine for about a year). I finally removed the daisy chain configuration, that I couldn’t be bothered to change, to connect a second PCIe cable, but 3DMark keeps shutting down the computer randomly.
Some Google searches have suggested that the PSU might be faulty. I’ve been using this PSU for about six years. It’s an Aerocool Templarius Imperator 850W. I calculated online whether the combination of new parts would require more power than what my PSU provides, but the calculation suggested I’d need about 630W.
So what might be the issue here? I’m considering the following:
1. Best case scenario, I’ve missed a drivers issue with the new Windows installation, although I have installed the latest nVidia drivers.
2. Might my GPU be at fault? Is there a way to test its components in isolation the same way MemTest86 or Cinebench do for the RAM and processor respectively?
3. Maybe my PSU has always been faulty, but I’ve only found that out now that the new processor, motherboard and RAM, added to the GPU, are stressing it more.
If I have to replace the PSU, I don’t look forward to dismantling half of my tower to remove the current one; it might be the hardest part to remove from my Dark Base Pro 900 tower once it’s installed.
I know I won't rest easy until I solve this problem, so I'm scouring online forums to figure out if I'll have to buy a new PSU and dismantle half of my setup, as I fear.
Intel Core i9-12900K 5.2 GHz
Asus ROG STRIX Z690-F GAMING WIFI
G.Skill Trident Z5 RGB DDR5 6000MHz 32GB 2x16GB CL40
Noctua NH-D15 chromax.black
The RAM sticks pass the built-in MemTest64, the processor withstands the Cinebench test, but when I tried to use my VR headset, the computer shut off. I downloaded 3DMark, suspecting there was a problem with the GPU (a GeForce RTX 2080 Ti that has been working fine for about a year). I finally removed the daisy chain configuration, that I couldn’t be bothered to change, to connect a second PCIe cable, but 3DMark keeps shutting down the computer randomly.
Some Google searches have suggested that the PSU might be faulty. I’ve been using this PSU for about six years. It’s an Aerocool Templarius Imperator 850W. I calculated online whether the combination of new parts would require more power than what my PSU provides, but the calculation suggested I’d need about 630W.
So what might be the issue here? I’m considering the following:
1. Best case scenario, I’ve missed a drivers issue with the new Windows installation, although I have installed the latest nVidia drivers.
2. Might my GPU be at fault? Is there a way to test its components in isolation the same way MemTest86 or Cinebench do for the RAM and processor respectively?
3. Maybe my PSU has always been faulty, but I’ve only found that out now that the new processor, motherboard and RAM, added to the GPU, are stressing it more.
If I have to replace the PSU, I don’t look forward to dismantling half of my tower to remove the current one; it might be the hardest part to remove from my Dark Base Pro 900 tower once it’s installed.
I know I won't rest easy until I solve this problem, so I'm scouring online forums to figure out if I'll have to buy a new PSU and dismantle half of my setup, as I fear.
date
newest »
newest »
message 1:
by
Shelby
(new)
Jan 01, 2022 02:54PM
Judging from what you've said, it sounds like a PSU issue. You can use FurMark or something similar to stress test your GPU. But with everything you've said, it seems most likely to be a PSU issue. Further googling leads me to believe that people using 3DMark who experience shutdowns most often find out the problem is attributed to PSU. Good luck.
reply
|
flag
Shelby wrote: "Judging from what you've said, it sounds like a PSU issue. You can use FurMark or something similar to stress test your GPU. But with everything you've said, it seems most likely to be a PSU issue...."The people that have answered a post I made on Tom's Hardware forums agree with you. My ten years old PSU that wasn't particularly reliable to begin with (it's the one they had available at the store when I went to fix my computer after a storm wrecked half of it back in 2014). I never experienced any issues with my previous motherboard, processor and RAM, but I guess my current components are far more demanding and need more stable voltages.
In any case, I bought a new PSU (full name: Corsair RMx Series RM850x 850W 80 Plus Gold Modular), but I'm not looking forward to taking my tower apart just to remove the previous PSU and install the new one. I'll probably have to remove the CPU cooler, reapply the thermal paste, etc. to reach some of the power supply cables connected to the motherboard as well.
In any case, thank you for taking the time to read the post and even offer advice.
I recently had to replace my PSU as well. I had a nice cable setup before, so I was pretty unhappy that I had to dismantle things to get it seated and then reorganize my cables. It's a nightmare when a part like this goes bad. I feel your pain. I hope it's not too bad in any case. At least PSUs come in modular now, they cut down on a lot of unnecessary cables. Take care, friend. :)
Shelby wrote: "I recently had to replace my PSU as well. I had a nice cable setup before, so I was pretty unhappy that I had to dismantle things to get it seated and then reorganize my cables. It's a nightmare wh..."I've gone through the tremendously annoying trouble of removing the PSU shroud from my Dark Base Pro 900, getting rid of the previous PSU from ten years ago and then installing my brand new Corsair RMx Series RM850x. I've run a full stress test on 3DMark, as well as a couple of regular benchmarks, and the system seems stable. I can't test more before I leave for work, but it does seem that the previous PSU was at fault.
Again, thank you for your help, and I hope things go well for you.
