Random PC shutdowns without BSOD
Random PC shutdowns without BSOD
I'm trying to determine what might be triggering my PC's sudden shutdowns. My system setup includes a Dell Optiplex 9020 with an i7-4790 processor, 16GB DDR3 memory, a 650W EVGA graphics card, a 500GB SSD, and a Zotac GTX 1070 Mini GPU. Recently, I upgraded the GPU from an EVGA GTX 750 Ti to a Zotac GTX 1070 Mini, but the issue started after that change. Since then, random shutdowns have occurred multiple times—three before I replaced the GPU, which I thought was a short. After that, another shutdown happened the previous day, and again last night. Each incident followed a full day of use, often ending just before gaming sessions. It doesn’t seem to depend on how intense the game is; it happens after long work hours and late-game sessions. Even with lower-demand games like Genshin Impact, it still occurs, sometimes during short playtime. I’ve tested my GPU, CPU, and PSU with various tools, and everything appears normal except for occasional spikes in power draw. The PSU is under warranty and has all the safety indicators intact. My case is open, but I suspect overheating could be involved—especially since the PSU shows no obvious faults. I ordered a PSU tester, but it feels like that might be the culprit. I’m worried about potential damage, even though the unit is new. Could there be another factor at play?
I possess the data but it only displays 6008. The prior shutdown occurred at 11:05:17 PM on 3/22/2023 and was not anticipated. Following that, a count of 41 appeared. The system restarted without a complete shutdown. This issue might stem from unresponsiveness, crashes, or sudden power loss.
it doesn't necessarily mean the PSU is overheating. I'd favor the motherboard VRM instead. these boards usually have 2 to 4 phases without a heatsink. the 750Ti performed well, but now you're more than doubling the heat inside the case with the 1070 (assuming you keep the side panel on, since you didn't specify otherwise). the i7 uses about 20W more than the i5, so if the issue was with the less powerful part, it will affect the upgraded model too. many factors can activate protections, and office towers often lack detailed sensor data. if cooling works, consider swapping the case or adding fans—two more. you could also create a larger opening for a 120mm exhaust. temperature issues are generally easy to resolve; you might even skip the side panel entirely. if needed, install a box fan directed at it...
I’m expecting the PSU tester tomorrow, which should help with some answers. Regarding the motherboard, you’re wondering if the extra heat from the 1070 Ti might be overheating the VRM and causing a shutdown for safety. I’ve noticed the airflow getting warmer when I touch the back of the PC while using the new card. Since neither the CPU nor GPU was overheating, I think everything should be fine—except for the possibility that the PSU is faulty, which I can’t verify. I’m considering adding a fan hub and adapter (though the ports here are proprietary) and possibly extra fans to see if it fixes things. My main concern is avoiding damage to other parts, especially the new GPU, during a power outage.
these safety features are designed to keep things safe, which is why it’s usually beneficial for them to activate rather than fail. I’d consider running with the side panel removed if you’re concerned about internal temperatures. (This test is quite simple.) Based on an average mATX build of 60L and roughly 200W (an i5 plus a 750Ti combo would be around 95W less), the power usage during gaming is minimal. With such limited airflow, the internal temperature would likely reach about 24°C after four hours, which is only a slight rise from ambient conditions. A solid PSU should handle this easily since it won’t be pushed beyond half its capacity. If temperatures were extremely high, it would mainly shorten the lifespan rather than cause shutdowns. With a 10-year warranty, this shouldn’t trigger protection. From what I see, the VRM might be close to its limits—we can’t see them clearly, so they’re probably hidden behind the EPS cable. I also recall that most DELL optiplex boards from that era used three-phase power, though I haven’t seen many in this setup. Building 16-phase motherboards is often over-engineered for marketing purposes; they usually operate well below capacity, typically staying around 70–80°C externally with a 150°C internal rating. So it seems like the most likely cause could be here.
The PSU tester didn’t pass the check. It wasn’t designed to verify performance under load, which is what you’d need for those expensive units. Still, it looks like you’re onto something. I’ve observed with the new GPU and CPU that the airflow from the back of the case has noticeably dropped. The 1070 Ti runs about 6-7 degrees cooler than the 750 Ti during short gaming sessions. When I was using HW64, the PC crashed and I checked the GPU temperatures—at shutdown they were rising, matching the CPU’s behavior. It’s likely the surrounding VRM area was overheating, possibly due to a weak fan or poor cooling setup. I tried running without the side panel using the 750 Ti, which kept temperatures around 6-8 degrees lower. I’m planning to buy some fans and a Noctua controller for better airflow and stability. This should help. I appreciate you breaking it down—previously I hadn’t thought about motherboard temperatures or overall room conditions.
Physics holds up well at this level, though the figures I mentioned are just rough estimates. They show the speed and heat buildup clearly. Fans usually aren't much help unless there are openings for airflow in a sealed setup. Some designs naturally reach higher temps. For VRMs specifically, my tests on my motherboard showed 80°C outside when the CPU was drawing 90W continuously. The MSI B350M board is budget-friendly... that’s why I chose it. Many thought these low-cost boards wouldn’t handle the newer AM4 CPUs coming out after 2017 because they were built for cheaper parts than those Intel boards. Most B350 models had 6-phase VRMs, which doubled to a 3-phase setup. Now it’s clear people overreacted; the engineering is solid. These very cheap boards can easily manage 160W if you direct airflow onto them. In short, your board is handling its workload similarly to mine, but with less power. Your cooler does move air over the VRM, which is a plus, though they lack heatsinks for extra surface area—so even with good airflow, it might struggle when things get hot.