The system is restarting randomly, yet it isn't overheating.
The system is restarting randomly, yet it isn't overheating.
I mentioned gold because I was checking a Corsair bronze piece, but BF thought gold is better since it’s more stable than bronze. However, as you pointed out, stability doesn’t always mean quality. I think I was looking at the RMx model (link provided) and wondered if 850w was really necessary, even though I know the PSU only needs the power it uses and going over isn’t a problem.
The PSU's performance is largely unrelated to its stability. There isn't really a concept of "stable" or "unstable" PSUs. What matters are the build quality aspects. From very poor (such as doorstop, fire risk, time bomb issues) to low, average, good, and excellent, there are different levels. Efficiency simply indicates how much power the PSU consumes from the wall and how much is lost as heat.
You can view the full chart here: https://en.wikipedia.org/wiki/80_Plus#Ef...ifications
Having more than the suggested wattage isn't excessive. However, going beyond that, for example to 1kW or higher, you'd face two drawbacks:
* Higher cost for a more powerful PSU
* Lower efficiency
The PSU works most effectively when the load is between 50% and 80% of its maximum rated power. For instance, with a 400W load on a 650W unit (like the Seasonic PRIME TX-650 80+ Titanium), you'd achieve about 61.5% efficiency, meaning only 4% of energy is lost as heat. In this case, the PSU would draw 416W from the wall and deliver 400W to the components, with just 16W wasted.
If you use a 1kW PSU (such as the Seasonic PRIME TX-1000 80+ Titanium), the efficiency would be around 94%, drawing 424W from the wall, delivering 400W to the load and losing only 24W as excess heat.
Conversely, with a 400W load on a 1kW PSU (like the Seasonic PRIME GX-1000 80+ Gold), efficiency would be about 90%, pulling in 440W from the wall, providing 400W to the components and wasting 40W as heat.
Overall, the RM850x is a better option compared to the RM750e because it comes with a 10-year warranty and superior capacitors, whereas the latter has a shorter warranty and lower-quality parts.
To confirm my understanding of the 2933 versus 3200MT/s comparison in an older system I recently refurbished.
From what I've learned, the AMD 2600X is officially compatible with DDR-2933.
I recently replaced an A6-9500 dual-core APU with a used 2600X and tested whether it could run with a spare DDR4-3200 instead of the DDR-2666 used with the A6-9500.
As expected, the CAS timing required minor adjustments to achieve stability at 3,200MT/s.
This machine is part of a collection of vintage computers I maintain for software testing. I enjoy trying new applications on such systems, regardless of potential issues.
Restore points don’t always function as intended when I need to revert changes. Hyper-V images are useful for evaluating fresh software, but certain programs require direct hardware access, making VMs unsuitable.
This reminds me of the challenges we face with regression testing new software releases at work—sometimes they fail and occasionally damage the testing environment.
I prefer that my 3800X and 7950X setups remain intact and unaltered.
Thanks for your support! This will be useful for comparison, and I plan to test it out this afternoon to evaluate the performance and decide on upgrading my PSU with the RMx model.
I might have solved it, but I don’t want to share too soon. The game kept crashing every ten minutes now after updating some recently released drivers (I managed to play for a few hours before that). In a final attempt, I turned off Ryzen Master’s core optimizer, which slightly lowered the CPU temperature (I like that for cooler runs). So far, no crashes have occurred. It seems the game is so demanding I needed maximum performance from my CPU. Temperatures reached low 70s (up to about 80 for a short time in certain areas), but everything looks okay so far.