Well I don’t know much about all this temperature stuff, I can study that more, but I really don’t think it’s a problem in my case. And about the input lag on CSGO it seems like it stopped, just like the BSOD and the problem with the fans. I’ve been using it for about a week now without any problems, I’ll keep using it to see if something happens, but if it’s fixed I don’t have ideia what was causing the problem. I was thinking maybe something related to my 1080Ti which has 2 energy cables connected to the power supply, and I was using just one, and know when I was changing the water cooler I decided to use another cable. So maybe that was causing the problem, I don’t know if this can or has anything to do with the input lag, but with the BSOD I’m 90% sure. Also I decided to keep AÍ Suite 3 to check the temperatures, but if I start having any problems will be the first thing I will check. But my friend used it for about 7 months without any problems...
Wow, okay. Uhm, that's the kind of thing you a) tell people about when troubleshooting, and b)
don't do. You're lucky (or perhaps not?) that the GPU would run at all with just a single power cable connected, the downside of that is that you've been running that cable (or the PCIe slot in your motherboard) significantly above its rated power delivery. A 1080 Ti is a 250W GPU. Thanks to GPU Boost 2.0 it will stick closely to that limit unless bottlenecked by something else. A single 8-pin PCIe power connector is rated for 150W (6-pin wires are 75W). The PCIe slot can also deliver ~75W, though that's across 3.3V and 12V, with ~65W the actual rating for its 12V rail. That means that with your current setup, if the GPU consumes 250W, one of the two
must be out of spec. Best case scenario, it draws the extra power from the power wiring, and the wiring is overbuilt enough for it to be an issue. But that's the best case scenario.
A middle-of-the-road scenario is power throttling, i.e. that the GPU is throttling down due to insufficient power delivery. This can be due to voltage droop, from the GPU sensing that too few power cables are connected, or other on-board sensors. This is especially likely with weak or insufficient wiring, as insufficient wiring will lead to cables warming up, which increases resistance and thus voltage drop over the cable, leading to the GPU getting a lower voltage delivered than it needs. This will cause the GPU to clock down to protect the VRMs and power delivery. It's not unlikely that this is your issue.
It's also worth noting that if your power cables have been running hot over time, the insulation will deteriorate and turn stiff and brittle. This can cause it to crack and create shorts. Check your cabling. It might be dangerous.
Just as an FIY, here are some of the possible worst-case scenarios from a setup overloading the power delivery:
-Burning out the 12V input in the PCIe slot of the GPU, likely killing the GPU
-Burning out the 12V delivery in the PCIe slot of the motherboard, likely killing the motherboard (and possibly taking the CPU with it)
-Melting the insulation on your PCIe wire or melting its connector, whether on the CPU or GPU side. Melted insulation leads to short circuits, which lead to either blown fuses (good-ish), killing your PSU (less good), or a full-blown fire (yes, that's bad).
Some motherboards have unnecessary power inputs for the CPU - any AM4 platform with more than an 8-pin EPS, for example, as getting Ryzen above even 200w is nearly impossible unless you're doing LN2 OC - but GPUs generally want and need all the power connectors on them. If you have an extreme edition GPU with three 8-pin PCIe power connectors but run it at stock it's likely fine with just two, but anything with two power connectors should have both connected. And they should each be getting power from separate cables from the PSU.
