Help a beginner - Screen turns on when games are running
Help a beginner - Screen turns on when games are running
My setup includes an Intel i7-4790k processor, a Corsair VS 650 power supply, a Gigabyte H81M-S2H motherboard, 16GB of RAM, a GTX 970 graphics card, and I usually manage to solve problems on my own with enough time. However, I’m running out of ideas. When playing older games, my display shuts off immediately. Right after, Windows will announce an error, and the audio continues playing through my speakers, even though I try to close Alt+F4 or Alt+Tab.
I’ve attempted various fixes—my CPU is overheating—and taken it to a repair shop to clean the old paste and replace the GPU. They also checked all components, especially fans, making sure they were properly secured. They recommended a new GPU, but it seems like they’re mainly trying to sell me one. I mentioned that icons appeared on my desktop after a clean Windows install, which led them to suggest a graphics card upgrade. When that happened, my Onedrive was downloading my desktop, so they insisted it was a serious display issue.
I also tried a fresh Windows installation and played Overwatch 2 with only the Blizzard app installed, but the problem persisted. I tested it with minimal settings (medium load at 60fps, then at 30 and lowest), and nothing changed. My system now runs only essential apps, and I’ve disabled most background processes on startup. I cleaned everything with wipes and dusters, locked Windows to 30 frames, and used Open Hardware Monitor to check the hardware. MSI Afterburner helped me test GPU performance, but it didn’t resolve the issue. I even ran the fan at maximum speed to try lowering CPU temperature, but it had no effect.
I suspect potential problems with the GPU, the power supply, or even the PSU’s capacity, though I’m aware of the overheating risk. The display issue is persistent, and I’m unsure how to fix it without replacing parts. In a tight budget, I’m willing to buy new components if they’ll solve the problem, but I can’t afford to replace everything just to test.
Did you buy it as prebuilt? 🤔 There is some truth in that. To test if your GPU is an issue, bench it, with Superposition, link: https://benchmark.unigine.com/superposition Use 1080p preset. Medium or High. Using Extreme preset will stress test your GPU beyond belief (you'd get only few FPS during bench). But still, Extreme preset is also good to be ran. Score wise, 1080p Medium preset should give you ~8400, 1080p High ~5900 and 1080p Extreme ~2100. If no artifacts appear during the bench and it finishes without issues, GPU is sound. This points towards GPU driver issue. Could be that you have a bad driver version. In this case, uninstall current drivers with DDU and install version or two older drivers, to see if it helps. Another thing to try would be trying the Win + Ctrl + Shift + B key combo once you're seeing the black screen. This command resets the GPU drivers and may bring the image back. <- This won't fix the underlying issue per se, but could save you from cold restart. This you need to replace regardless. Depending which version you have, your PSU is either low quality (gray label) or complete crap (orange label). With PSU, it is very possible that over the years, it has fed out of specs voltages to your hardware, damaging them over time, until to the point that you'll see issues or complete hardware failure. Bad PSU is also one of the main reasons why PC can't operate during heavier load (usually gaming). For new PSU, get good/great quality unit. E.g Seasonic Focus/PRIME or Corsair RMx/RMi/HX/HXi/AX/AXi. (All 3 of my PCs are also powered by Seasonic, full specs with pics in my sig.) For 2nd opinion about your current Corsair unit, look it up from PSU tier list, link: https://forums. It is either Tier D (gray label) or Tier E (orange label), while what i suggest and am personally using, is anything from Tier A.
Initially, thank you very much for your time and the details you provided.
Yes, it was about 8 or 9 years ago and has been fine until this year.
Recently, I performed a test and everything worked smoothly at medium 1080p. Then I tried high resolution and it caused my display to crash. I attempted to restart the system and use the benchmark tool immediately afterward, which showed improved performance without issues. The FPS remained steady at 60 on medium, dropping to 40-45 on medium and 15 on extreme.
This behavior is common—I often experience crashes within five minutes when playing a game for the first time. After restarting, it would work well for an hour or two before failing again. Changing the game settings doesn’t seem to help.
I have an image I took for you; it looks like a poor one (https://ibb.co/vsMXkPm). I’m not entirely sure about hardware compatibility, though. I know the components and their models, but I haven’t assembled a system before, so I’m unsure which parts are suitable. I attempted to estimate the power requirements, but there’s no clear guide available.
Considering this, I think a new PSU and GPU might be necessary.
It’s clear from the choices made that the build was preassembled. The top-tier CPU paired with a barebones motherboard is evident. Above all, using a K-series CPU alongside a Z-series motherboard—such as the Z97 chipset—would allow for CPU overclocking. It’s also worth noting the consideration of an Intel 80-series chipset versus a 90-series one. The latter was designed for Haswell Refresh CPUs (Core i5/i7 K-series), though the 90-series is still functional but requires the latest BIOS. Additionally, opting for an Intel 80-series chipset would be problematic, as it wasn’t built for that generation.
The selection of a PSU also raises concerns; many components were incompatible at the time. It seems the power supply was outdated and should be upgraded immediately, particularly given its age. For better compatibility, tools like PCpartpicker can help—link: https://pcpartpicker.com/list/. This platform prevents incorrect selections, ensuring only suitable parts are chosen.
For CPU cooler compatibility, for example, picking an NH-D15 would restrict cases with less than 165mm clearance, as the cooler wouldn’t fit otherwise.
Regarding power consumption, there are calculators available that estimate the right wattage based on your PC specs. I use manual methods instead: checking GPU TDP (e.g., GTX 970 at 185W), accounting for transient spikes with RTX models, CPU TDP (Core i7 at 88W), and adding ~150W for other components. This totals around 423W, plus a safety margin of ~200W, resulting in roughly 623W—suggesting a 650W PSU is sufficient.
When choosing components, especially PSUs, it’s wise to consult reviews from trusted sources. If no reviews exist, avoid the purchase. Since the PSU powers everything, it stands out as the most critical part.
Reading expert opinions can be time-consuming, so seeking insights from professionals like TH experts is beneficial.
Upgrading to a new PSU is essential. As for the latest GPU, I’d first replace the PSU and test it with it. It’s possible your current PSU is the issue, allowing you to extract more performance from your GTX 970. If problems persist, testing with a second GPU (borrowing one) or removing the GTX 970 entirely to use the CPU’s integrated graphics might help.
However, if you decide to upgrade the GPU, ensure your PSU matches its power requirements—especially since GPU power draw has increased significantly. The era where a 650W PSU was adequate for any single GPU is behind. Currently, the RTX 4090 requires around 600W under load (with spikes), needing at least a 1600W unit. A 650W PSU may not suffice for top-tier GPUs like the RTX 4090.
If your new GPU is chosen, assess your PSU’s capacity based on your budget or opt for a 650W model with a clear understanding of its limitations.
You previously checked a power consumption calculator and determined you required between 400-450 watts, which is reassuring. However, I wasn't confident the information was accurate or that I had interpreted it correctly, as I suspected my PSU might be of poor quality. I was also searching for the optimal GPU for a budget card, ultimately settling on the GeForce RTX 3060. Still unsure if it would fit in my motherboard or case, so I left it there. Recently, I explored options and found that a PCI Express 4.0 is necessary for the card, which should match your motherboard model. If that’s the case, my current motherboard supports only PCIx1 and PCIx16 slots. I wondered whether a higher slot could accommodate a lower-numbered PCIe version. Could a PCIx4 slot fit a PCIx16 one?
I also performed some calculations based on your previous numbers. You mentioned the GPU uses 88 watts, the rest 150 watts, and the RTX3060 needs 170 watts. Adding them gives 408 watts plus a buffer of 200 for safety, totaling 608 watts. A 650-watt PSU would still be sufficient. I was considering buying more power just to ensure stability, but based on our discussion, it might be better to opt for a lower wattage but higher-quality PSU for the price. I was planning to spend around £100 (about $125) on a good PSU, which feels like a reasonable investment before my payday next month.
If this solution resolves my concerns, I won’t worry about a new GPU just yet. However, I’m eager to upgrade to a better one in the future and would appreciate recommendations for a reliable PSU tailored to my needs. I’ve noticed most reviews are overly positive, so I suspect some bias. I found a Silverstone SST-ST70F-ES230 700W model, but I’m unsure if it’s worth it. The color option listed is black/gold—gold seems like a decent choice?
PSU quality is actually far greater aspect than the PSU wattage. Sure, wattage defines how beefy hardware you can run, but quality defines how long you can use your PC, be it days/months (with crap quality PSU) or easy 10+ years (with great quality PSU). Also, the worse the quality of a PSU - the higher the chance PSU failing and also the higher the chance of failed PSU frying everything it is connected to. Heck, there have even been cases where failed PSU has catched fire and burnt down entire homes. Even lives have been lost.
I see that you've mixed up the PCI-E a bit.
With PCI-E, there are two different things;
* revision
* slot size
Revision is to do with bandwidth and is written in from of: PCI-E 1.0, PCI-E 2.0, PCI-E 3.0, PCI-E 4.0 and the latest PCI-E 5.0.
Your MoBo has PCI-E 2.0,
specs:
https://www.gigabyte.com/Motherboard/GA-...v-10/sp#sp
PCI-E revisions are backwards compatible. So, you can use PCI-E 3.0 or 4.0 GPU in your MoBo. But it would work with reduced bandwidth, since every PCI-E revision doubles the bandwidth.
Slot size is written as how many lanes the slot has. E.g PCI-E x1, PCI-E x4, PCI-E x8 or PCI-E x16.
This image shows nicely the PCI-E slot sizes;
All GPUs have PCI-E x16 slot and only fit into x16 slot. This is even true when GPU only uses 8 lanes (PCI-E x8), but it still has the physical PCI-E x16 slot. One such example would be AMD Radeon RX 7600, which uses 8 PCI-E lanes but has 16-lane slot,
specs:
https://www.techpowerup.com/gpu-specs/ra...7600.c4153
RTX 3060,
specs:
https://www.techpowerup.com/gpu-specs/ge...2-gb.c3682
is PCI-E 4.0 x16. Meaning it uses PCI-E revision 4.0 protocol and 16-lane slot.
Also, you've typed "PCIx16", which actually refers to the old, PCI protocol, predecessor of PCI-E. On the image above, you see two PCI x16 slots as well. PCI is obsolete nowadays.
When it comes to PCI-E slots, other add-in cards, e.g sound card, is usually PCI-E x1. But you can plug them into longer slot as well and they work in there too. But you can't physically plug longer slot device (e.g GPU with x16) into smaller slot (e.g into x8).
PCI-E x8 slot is very rare and hardly ever used on today's MoBos. Same goes to PCI-E x4 slot as well. Nowadays, MoBos have PCI-E x16 and PCI-E x1 slots. E.g even your MoBo has one PCI-E x16 and two PCI-E x1 slots.
Yes. For RTX 3060, 650W unit is enough.
Well, you can, but there are two cons;
#1 Higher wattage PSU costs more than lower wattage PSU.
#2 Loss of efficiency.
To talk about #2 more; PSUs are most efficient when load on them is 50% - 80% of their total wattage.
408W load on 650W PSU would mean 62.7% load on PSU.
Now, if you'd have 1kW (1000W) PSU, the load on PSU would be ~40%, making PSU itself less efficient. And efficiency translates into how much power PSU pulls from the wall and in turn, how much of it is wasted as excess heat.
To explain PSU's efficiency, they are rated by
80 Plus standard
; whereby different levels are;
* 80+ (aka White) <- Never buy 80+ PSU since at current date, those are low/crap quality.
* 80+ Bronze <- Used to be the norm ~15 years ago. Nowadays, best to be avoided.
* 80+ Silver <- Only very few PSUs had it.
* 80+ Gold <- Current norm.
* 80+ Platinum <- Only few PSUs have it.
* 80+ Titanium <- Best there is.
Here, in efficiency calculation, i'll take my own PSU, Seasonic PRIME TX-650 80+ Titanium [SSR-650TD] as an example (btw, the best PSU money could buy at the time of purchase, back in 2016, and still, one of the best, if not the best, PSU out there).
80+ Titanium PSU has efficiency rating of:
On 20% load - 94%
On 50% load - 96%
On 100% load - 94%
So, for e.g. 400W load on my 650W unit, would be 61.5%, meaning that the PSU is 96% efficient, where only 4% of power is wasted as excess heat. Meaning that the PSU draws 416W from the wall, gives 400W to components and wastes only 16W as excess heat.
Same 400W load on 1kW PSU (e.g Seasonic PRIME TX-1000 80+ Titanium) would mean that the PSU is 94% efficient, since load on PSU would be 40%. Meaning that PSU draws 424W from the wall, gives 400W to components and wastes 24W as excess heat.
But if you'd get 1kW 80+ Gold PSU, which is the norm today (e.g Seasonic PRIME GX-1000 80+ Gold), with efficiency ratings of:
On 20% load - 90%
On 50% load - 92%
On 100% load - 89%
Would mean that on 400W load, PSU is 90% efficient. In this case, PSU pulls 440W from wall, gives 400W to components and wastes 40W as excess heat.
So, would you rather have 650W 80+ Titanium PSU, which only wastes 16W as excess heat on 400W load? Or 1kW 80+ Gold PSU, which on 400W load, wastes 40W as excess heat?
Also, while the excess heat waste numbers are small, keep in mind that they add up over the time. E.g if time advances tenfold, the excess heat numbers would be 160W vs 400W, with a diff of 240W.
Of course, 80+ Titanium PSUs cost a lot. My PSU costed me €206.80, but since i also have UPSes in use (both on my and my missus'es PCs), i can't afford loosing UPS'es runtime due to PSU's inefficiency. That, and it also adds up to electricity bill (electricity is expensive nowadays).
I live in EU (Estonia). Though, 100 quid is a bit less for proper PSU. But you could get Seasonic Focus GX-650 for it,
pcpp:
https://uk.pcpartpicker.com/products/com...TW,TN3mP6/
I also added some other Seasonic units into pcpp link (all are Tier A), including my PRIME TX-650, so you can see how much the best costs.
Seasonic Focus is good quality PSU with nice 10 years of warranty, among other things. Even i have Seasonic Focus in use, namely Seasonic Focus PX-550 80+ Platinum, powering my old AMD build (full specs with pics in my sig), while our main builds, both mine and missus'es, are powered by Seasonic PRIME TX-650 80+ Titanium PSUs.
Oh, if you go with Seasonic PRIME (any efficiency), then you get the best made by Seasonic, with mythical performance and 12 years of warranty (longest there is).
Looking the PSU by it's color?
🤣
It's like thinking the RGB LEDs (aka unicorn puke) in the PC adds more FPS to games.
😆
Silverstone used to be good, ~20 years ago. Nowadays, Silverstone mainly makes poor PSUs. But there are some good ones as well.
That specific Silverstone unit (SST-ST70F-ES230) is low quality (Tier D) unit. Sure, better than your Corsair VS but nowhere near the level of what should be used in a PC.
Refer to the PSU Tier list i linked earlier, to know which PSUs are good and which aren't.
For any PC that has dedicated GPU, only Tier A PSU would do. If you'd have low-spec PC or office PC, then Tier B would do too. Other, lower Tiers, are best to be avoided.