The game becomes unresponsive while sound continues to play.
The game becomes unresponsive while sound continues to play.
Hello everyone,
I'm facing a bunch of issues with my PC that are proving difficult to pinpoint. With over two decades of experience assembling computers, this one is particularly puzzling since it's nearing its 5th anniversary now.
Here are the details:
- CPU: Ryzen 7 3700X
- CPU cooler: Noctua NH-U12S
- Motherboard: MSI b550 Tomahawk (MS-7C91)
- RAM: 4 x 8GB Corsair LPX 3200 XMP 2.0
- GPU: Gigabyte GTX 970 Mini-ITX OC 4GB
- Power supply: Corsair RM650x
- SSD: Samsung 970 EVO Plus 1TB
- Chassis: Fractal Define 7
- Operating system: Windows 10
- Monitor: iiyama g-master g2530hsu
The GPU is a second-hand unit originally, as I initially planned to fit a 3070 model. However, due to the post-COVID price surge, it became more affordable, and it has performed reliably ever since.
BIOS settings:
- Version: 7C91vAJ
- CPU core voltage: 1.300 V (instead of 1.500 V in [Auto])
- XMP: Enabled (problems were tested with both enabled and disabled)
- PBO: Disabled
Everything seemed normal until about 4–5 months ago, when the issues started. At first, it was uncommon, so I ignored it until it worsened.
The symptoms:
- The screen freezes (green, orange, or black), audio keeps playing, and I have to restart by holding the power button for 5 seconds.
- After restarting, the PC restarts without issues.
This occurs during gameplay, though not all games do. ArcheAge has been my go-to test title because it crashes within 2 minutes once past the selection screen.
I quickly suspected the GPU. Here’s what I tried to fix it:
1) Performed a thorough cleaning of my GTX 970—disassembling fans, heatsink, and reapplying thermal paste. I took my time.
2) Installed all official Nvidia drivers one by one using DDU each time.
3) Reset the CMOS—no improvement.
4) Connected the 970 to my girlfriend’s PC and her RTX 3070. The test game crashed in about two minutes on her machine, just like before.
5) After an hour of normal operation with her card, I ran another test.
Key observations:
- These crashes happen before the GPU or CPU is under load; it's not a thermal issue. All components are cold during the events.
- The problem doesn’t occur when I’m under stress.
After this, I decided to purchase a new GPU—MSI RTX 5060 2x OC on Amazon at a reasonable price.
5) I received the new GPU, installed it correctly, booted into safe mode with no network for DDU, updated the latest official Nvidia drivers. The test game ran smoothly immediately.
Now comes the second issue:
- It lasts about 5 minutes before freezing, showing a black screen and background audio. I manually shut down the PC.
- EXCEPT this time, starting my computer was impossible:
- No boot,
- No video signal,
- GPU fans running at full speed!
I looked up possible causes and tried these steps:
6) Reset CMOS – no effect.
7) Checked 5060 connections – still no fix.
8) Installed 5060 on my girlfriend’s PC. She booted the game, but it crashed again in about two minutes.
9) Tried resetting the motherboard beep (3 long, 3 short beeps – HP style).
Checked connections multiple times, as the beep suggested an issue, but nothing resolved it.
10) Reinstalled the GPU and restarted my system. It worked perfectly afterward.
I also noted that these crashes happen before load; no overheating detected. The components were cold during the incidents. Also, no smoke or damage was visible on the 5060.
Eventually, I returned the 5060 to Amazon for a refund without a replacement, as my confidence in the decision wavered.
Now I’m unsure what to do next—especially since I’m trying not to buy unnecessary parts. I’m currently researching forums and Reddit to gather more insights before making any decisions. My main goal is to resolve this efficiently.
To sum up: I need to replace my GTX 970, but also wonder if the new 5060 from Amazon was faulty. Could it be a motherboard or PSU issue? I’m also concerned about damaging the brand-new GPU. Should I consider a PSU upgrade, a new motherboard, or is this just another repair?
What should I do next?
The likelihood of the new card being defective is minimal, quite so. You didn’t specify the model and brand of your system in the details. What performance levels do you expect for both the CPU and GPU when running at full load and when idle? For the PSU: Corsair RM650x
The case is a Fractal Define 7 (I included this detail in my main post).
This PSU is five years old, just like the rest of the equipment. The only exception is the GTX 970, which is second-hand.
It has only powered the GTX 970 so far, except during tests with my girlfriend's RTX 3070.
For temperature testing, I ran a 10-minute stress test on GPU and CPU together at https://mprep.info/gpu/, but I couldn’t maintain it for longer in regular gaming sessions. From what I remember, the temperatures stayed within normal ranges before then. (IDLE at minimum settings on simple browser use)
I dont believe anyone will ever have the exact same scenario. But here the update of my problem:
The RTX was just DOA.....
DOA -> overthinking the problem and looking for another solution when it was just the initial thought
My GPU was dead, nothing else !