Edgar R. Zakarian

Pretty impressive, thanks! I decided to do with Corsair RMX 1000W Gold, just to be on the safe side. My sis lives like 300km away, so I'm not gonna be troubleshooting for her. Better be on the safe side, besides costwise they were almost the same here. 850W vs 1000W

I saw this issue once. Problem then was windows 10 missing updates. After updating Windows 10 fully, problems disappeared. Not sure what exactly was the problem but I just found that solution on another forum at that point.

Is there a similar list somewhere for the newer AMD 6xxx or RTX 3xxx? Could this list be updated to include those maybe?

I've had this issue. Strange enough the GPU wouldn't work in another PC either O.O. Then I put it into the original PC; reset CMOS, turn off power to the PC for 10 minutes, booted it up and everything worked just fine - Also in the other PC. This was done on a GTX 680. Try resetting CMOS! It sometimes works.

I found some other posts as well which suggest it could be bad capacitors on PSU or the motherboard, so not exactly sure what's going on Anyone ever heard of a computer not starting because the rooms too cold? | TechPowerUp Forums

Experienced this. PSU's (Power Supply) have a weakness for cold and humid environment. I assume it starts working with your airconditioner because it dries out the PSU. I had an old PC where I had to blowdry / heat the PSU before starting, or else it simply wouldn't power on! This is typically caused by too cold storage, if your PSU has been stored at near 0°C pr similar. Definitely sounds like PSU defect.

Also try to specify which benchmarks you're exactly experiencing worse performance in, so we have an idea of what it could be. "Benchmarks" In general is too open. Is it a specific benchmark like Cinebench? Time Spy Extreme? Is it a specific game at specific settings? What resolution?

I had this problem with GTX 580. Never really solved it. But try enabling "disable ULPS" in MSI afterburner. Ultra Low Power Saving I think it means. I think last time I chalked it up to defect GPU. Something about it running in 2D mode instead of 3D mode. - Essentially downclocking itself to power saving mode, even though it's at the wrong time. Perhaps older drivers could work / remove the issue? Try 2 years earlier drivers or so.

Try Display Driver Uninstaller and reinstalling drivers from scratch. My PC used to do this with and old GPU, turned out to be a graphics card driver problem. In my case it was freezing my whole PC to a point where all I could do was hold down the power button and kill it. Reset button wouldn't even work lol. Also try re-seating the GPU. (Taking it out / putting in again). Re-seating the ram. Just making sure everything has proper contact. could be small shakes/vibrations cause the on-the-edge-contact to loose contact in some cases.

You mentioned watercooled etc. etc. But did you actually check the temps on CPU and GPU? Are they actually cool and running at good clocks? If your cpu was underclocking/thermal throttling itself it would explain bad FPS regardless of the GPU.

HAH! I think you're running into the same problem I am. Power Spikes from the RTX 3090 on a 750W PSU. I have the exact same setup, and in some demanding games, especially with Ray Tracing, the PSU just gives out, power goes out and PC restarts.

Oh no, not at all. Actually I experienced the problem AFTER changing the cables. That's because I only bought Dying Light 2 after changing. So I think it's not related, or the adapters could've possibly been worse So using the adapters from SATA to PCI-E I tested Time Spy Extreme, DOTA2, CSGO, Elden Ring (as previously pointed out is capped at 60FPS and has no Ray Tracing. So maybe I've just not been fully utilizing the card before Dying Light 2 on ULTRA. DLSS/SCALING didn't remove the crash problem either So possibly something to do with the RT Cores.

No idea, it worked great, but never tried Dying Light 2 on it like that hah. Prob. a good thing it didn't last long. Don't understand your question

Will do! I have no other options, so it's connected via 3x8-pin connectors today as well. Truth be told, I had previously overseen that my PSU has 4x6+2 pins. So I had used 2x6+2pins and the last one was converted from Sata -> Molex->2x6pins PCI-E->1x8pin PCI-E I ran Benchmarks like this, worked like a charm. But of course when I noticed i had the other cables to spare, I replaced all the janky adapters

So maybe I'm just cutting it waaaay too close with a 10 years old 750W psu xD

As I mentioned above, it's a temporary card regarding another post. Please see point 2 below.

Simple! My screen is that resolution & 144Hz

Have you tried overclocking ? Are you on the edge?

Title says it all. Please recommend some good ones which you think will handle the system then more. Links for europe if possible plz. (Preferably Denmark )

Huh you could be right. It seems like Raytracing / running the game on ultra is causing it. 7600k system is what I have, I bought the 3090 (used) as a gift and am just enjoying it until I pass it on ! Think 1000W unit will be enough for eventual power spikes? Link on deal? Europe?

To answer your questions: 1. Underclocking (just reducing clocks) definitely helped. I think it also reduced the wattage to about 350 as well. 2. It's actually doing great for the games I'm playing, even though it's pushing the CPU to 100% Definitely. The purpose of the card is actually to be given a gift to my sister who does VR Development in Unreal Engine 5. So calm your soul. It won't be wasted for 1080p gaming I just bought it used for a good price, and am enjoying it until I give it to her. It should very probable, could be what's going on. Just wondering how high of a wattage I need to get her, in order for her to fully utilize the card

My CPU is definitely bottlenecking it. Especially in other demanding games. But for 1080p Dying Light 2, it's doing fairly well. Oddly enough the crashes only seem to happen when Raytracing is on. If it's off, the game runs fine on ultra 99% usage, 2000MHz no problems. So yeah 1080p Ultra, 120-140 fps is prob. pushing the card somewhat.

Specs Intel i5-7600K (MSI motherboard overclocks to 4GHz automatically) MSI H110 PC MATE 2x8GB Crucial Ballistix LT Sport 2400MHz RTX 3090 Palit GAMEROCK Dark Base Pro 900 (Orange) Toshiba 250GB 2.5" Drive - Game Drive Crucial MX200 250GB 2.5" SSD - Boot Drive PSU Cooler master V750 (750W) semimodular Temperatures are around 72°C but when running Dying Light 2 at stock configuration, I can see that the card is drawing around 420W of power, after some minutes the whole PC suddenly shuts off and restarts. This is the PSU Current protection triggering, right? Strange part is, other games like DOTA 2, CSGO, Elden RIng or simulations like Time Spy Extreme have not done this You guys think it's just the game causing major spikes? Note: I noticed it shuts off especially when it runs around 2000MHz. Turning down the sliders for clock speed (-500) and memory speed solved the problem. (-500) Strange part is, limiting the power draw to 75% or 50% wasn't enough, it would still boost to 2000MHz and die. What do you guys think would be a fair upgrade for cpu to prevent this? - 1000W? 1200W? just to "futureproof" a bit too.

Makes sense! Thanks for the explanation! I guess higher heat -> 20°C -> 90°C is a higher gap of thermal cycling, thus worse for the "expansion" of materials and microcracks, maybe that's what I've remembered it from.

Isn't the lifetime usually defined and tested by x number of operation hours at y temperate ? That's why higher temperature shortens lifespan of electronics ?