I think the issue might be with the power supply, but let me know your thoughts before proceeding.
I think the issue might be with the power supply, but let me know your thoughts before proceeding.
Hello everyone! Hope everything is okay. I’m trying to understand what’s going on with my PC lately. Since January, it keeps shutting down whenever I try to play games or edit videos—especially when using Nicehash. It’s like it’s stuck in a loop: the fans spin, the DRAM and CPU lights blink, sometimes the GPU freezes with a black screen, and other times it just won’t power on at all. A power cycle helps, but even then, I can connect via TeamViewer and restart from there. It doesn’t reboot itself; it stays in those odd states until I turn it off again.
At first, I thought it was overheating, but after replacing the thermal pads, technicians checked and found everything normal. They even left it at their store for a weekend test, and temperatures stayed under 70°C—still okay in a room without much airflow. The card itself seemed fine.
Software-wise, I’ve got the latest updates from Windows 11 and AMD. But there’s a problem: on one of my drives, I have an older Windows 10 installation that just powers off like the newer one. DDU didn’t work either. As for Linux, I haven’t seen any shutdowns yet, but I’m still trying to figure it out.
I’ve been looking into possible causes—blown capacitors, faulty thermal paste, or maybe something else. I disabled Precision Boost Overdrive and XMP settings because of compatibility issues. The RAM sticks I have are not working as expected either. There was an incident in January when the GPU cable connector melted in the PSU, causing more frequent shutdowns and a white LED on the GPU saying “I don’t have power.” Some people warned me about overclocking or mining, but I didn’t do that.
After watching Linus recommend Nicehash, I started mining without issues until January. My settings were fine then, but now they’re not matching what AMD suggests for my Radeon. The fans are loud, and temps never go above 60°C—unless I’m in a room with a really noisy fan.
My PSU is a Cooler Master V700, which is modular and supposedly more efficient. I bought it used on eBay, so its age isn’t clear. It supports multiple PC builds, which is good for me. But I’m worried about missing something else to check.
I need advice: what should I do next? Should I get a second opinion before spending money on a new PSU? Maybe test the power supply with a friend. The years it’s been with me suggest it might be old, but I want to be sure.
My setup is:
- Gigabyte B450 Aorus Pro WiFi
- BIOS F62c
- AMD Ryzen 7 2700X
- 24GB DDR4 RAM (2x4GB Crucial 2400MHz + 2x8GB G.Skill 3600MHz)
- Radeon RX 570 8GB (Gigabyte)
- Windows 11 SSD, Windows 10 HDD, Ubuntu 22.04 LTS on 1TB HDD
- No other unusual hardware issues
I’m trying to troubleshoot this odd behavior and want to avoid wasting more time or money. Thanks for your help!
It’s a good idea to include a Tl;dr in this post, it helps grasp the main points quickly. All the details are welcome, but having a brief overview makes it easier to revisit specifics later. In short, from what I understand (mostly a quick read), your system is shutting down completely but still active. This might mean several things. My initial concern would be a faulty GPU or drivers, though diagnosing a serious GPU issue usually involves swapping it out and testing. That’s not very insightful. It’s more about efficiency—at full power (~50%), the system uses about 9W extra. That’s roughly the same as leaving a bright LED bulb on all day. If you run it non-stop at that level, you’d pay around $1 extra each month, depending on your electricity rate. Considering the cost gap between Gold and Platinum units, it would take over three years to recoup that expense. I wouldn’t recommend using PSUs under 80W, mainly because better efficiency matters more than price alone. A higher-quality 80+ bronze unit is likely more efficient than a gold one, so upgrading would be sensible. Still, I’d probably go ahead and replace the PSU, especially since it’s around eight years old—better to upgrade now rather than risking performance later.
From your description it seems the graphics card is the issue. That’s unusual. The RX570 consumes relatively little power, so it wouldn’t normally cause the PSU cable to overheat unless there was a fault such as a short circuit on the card, a faulty MOSFET or resistor, etc. The V700 was launched in 2013, so if you purchased it in 2014 it would have been fairly new at that time. It received an 80+ Gold certification during testing, which makes me wonder what they meant by platinum. Cooler Master offers a different “V” series model that meets 80+ Platinum standards, but these are newer and not the same as yours. It’s possible the person was mixing up your V series PSU with the more recent 80+ Platinum version. Unless they were simply advising you to upgrade to an 80+ Platinum power supply for some reason, I doubt they fully understood the situation. Based on the symptoms you’ve shared, I’d attribute the problem to the graphics card. However, since a cable melted and it’s been around 8–9 years, replacing the PSU might also be wise.
Absolutely on point, I appreciate it. This time things are definitely going wrong, and summarizing everything in a quick summary is tough. Still, I’ll think about it next time. Sorry for the long explanation—I was just trying to wrap it all up. The person mentioned was JayzTwoCents, and Jonny Guru from Cooler Master also talked about similar things before. I guess having an 80 Plus PSU helps gauge how solid the components are, which affects their lifespan beyond just the warranty. I picked the V700 mainly because of good reviews and Cooler Master's reputation at the time (Linus used those too). Both articles are pretty old compared to the PSU itself; they’re not the latest V series with higher wattage. It’s a bit concerning since the cables seem to be from years ago, but I hope it’s not the GPU—fixing the PSU might solve the issue. I’m considering bringing both to my repair person. Thanks for the response!
I hope so. The melted cable was included with the PSU. The replacement was also in the box—it's the final one. The motherboard BIOS is F62c (not the newest, but it's raining and I don’t have the UPS. It’s a risky situation, even with dual BIOS). The GPU VBIOS is 015.050.002.001.000000.
It's not. The connection lies in the PSU's efficiency and quality. If you invest in strong power delivery, it makes sense to build it efficiently too. Creating a PSU with poor performance or safety risks is straightforward, but achieving high efficiency can be simple by selecting a few reliable parts.
Absolutely, you're on the right track. I always check reviews to make sure. The tech staff I spoke to today recommended mostly EVGA bronze power supplies under $70. I wasn't too impressed by the fact they weren't fully modular like the V700. Most people told me those units could support my components for years. It was odd when someone suggested a lesser-known brand's PSU—though not extremely obscure in my area—but without proper certification. Charging a $100 unit with non-certified, non-modular cables really caught my attention.
Start with the fundamentals. I mostly read the thread—it brought some humor, but it was serious in the end. The board is budget-friendly, not ideal for a top-of-the-line flagship. Mixing memory issues with application errors isn’t unusual. If your PSU is nearly ten years old and you’ve already looked into it, it’s time to swap it out. Don’t care about other ratings on an older unit; just get a fresh one. Regarding storage drives, especially the green ones, they’re energy-efficient. They usually power down when idle, but take a while to start up and fetch data. I recommend keeping them in active mode and avoiding letting them rest. This should reduce any complications.