Check if the CPU overheats or if the power supply is struggling to meet demand.
Check if the CPU overheats or if the power supply is struggling to meet demand.
Mainly using a standard PC, some minor improvements expected: Power down during unexpected high usage (e.g., new small games like Dune Spice Wars). Loaded a single-player game at high settings, then restarted after powering off. Performance improved on low settings, with smooth operation and no crashes. Games ran well without sudden shutdowns.
Starship troopers had power-off moments near the end of matches. I used NVIDIA presets to cap the frame rate at 120 FPS. This helped avoid crashes going forward. It seems these games are not fully optimized, but I haven’t found a way to lower settings in other titles like Heroes of the Storm. Reducing sound and quality fixed some stutters, though low graphics still caused pauses every few seconds.
I’m trying to fix occasional ticks that keep appearing. Since I upgraded my motherboard, RAM, and CPU recently, I’m hoping these changes help. The issue appears to be more about setup than the game itself.
Current troubleshooting: Stress tests didn’t trigger shutdowns; CPU temps seem critical. I added case fans and swapped in a Noctua NH-D15 cooler (82.5CFM). Now I’m considering thermal paste and possibly a new CPU cooler.
Power supply is still solid – an EVGA SuperNOVA 750 G2 (80+ Gold) has been reliable for 8 years. With the new CPU and board, I’m aiming for at least 750W to support RTX 3080. There’s a risk of performance drop if it reaches limits during load.
Recent changes: Removed the old hard drive rack, opened airflow around the GPU, and cleared blockages on the CPU side. Top area is clear, but front and back have disk access.
Budget: Around $250. Should I upgrade the power supply or look for a better cooler? What kind of CPU cooler should I install? I’m confident my fans are working hard.
Liquid cooling could help more heat removal. Is the PSU upgrading enough to handle the new components? The 12V rail dropped from 12.2V to 11.8V during stress tests – it’s close, but worth checking if it affects stability.
I’ve removed the existing hard drive and improved airflow by opening the path to the GPU. Front side is clear, but IO and disk access are still blocked.
It seems there might be some problems on the way. Most likely, the 3080 spiking and PSU aren't handling it well. The NHD15 works just fine if you lower the voltage on the 12700k. Since you have a Z690, it’s probably because the system is running at peak performance, which can cause excessive heat and voltage spikes.
I'll try that to see if it stays stable around 80 degrees Celsius and performs consistently, which should fix the stuttering issue. Thanks for checking again, Jason. The new power supplies look promising—I'm hoping they’ll last over five years: I guess it’s my brand loyalty if I’m sticking with gold-rated options around $150 USD. For example, the EVGA SuperNOVA 1000 GT and MSI A1000G both have solid reviews. The plan is to spend about $260 USD on a PC that lasts roughly ten years.
An 850 is sufficient, but for 150 it's quite impressive to have an Evga model.
Refresh your profile details then recheck. https://www.msi.com/Motherboard/MPG-Z690...FI/support
The latest BIOS update includes improvements for CPU voltage control and highlights a prominent question about whether the CPU is boxed, tower-style, or watercooled. I already set this to tower cooling. Nice they made it a big popup now. I plan to run in Mode 8—OV isn’t available there at least. I’ll check if stability improves in-game and try to trigger power-off states during Dune Spice wars. The lower voltage setting should help with the new BIOS. New BIOS: Old BIOS: Realbench stresstest with Mode 8 selected, CPU voltage variation is reduced now:
Operating with CPU Loadline Calibration Control Mode 8, reviewed Dune Spice wars, still leading to power out issues. Video attached: textures set low allows the game to function properly. Possibly the CPU overheats and shuts down quickly, or the PSU is drawing too much power causing a crash—hard to say for sure.
Isn't your PSU covered by a ten-year warranty? It could be useful to reach out to EVGA and find out if they've encountered similar issues with 3080s. The 3080 is definitely a hassle, as my unit works well on a Corsair RM650x but fails unexpectedly on a Cooler Master 650W. I think it might be due to a voltage spike pattern that an older PSU wasn't designed to handle.