Thermal throttling on the CPU happens when it reduces performance to prevent overheating.
Thermal throttling on the CPU happens when it reduces performance to prevent overheating.
Hi All, Here are all my specs: MOBO: Asus m5A78L-M/USB3 CPU: AMD FX-8320 RAM: 16GB Kingston Fury GPU: XFX R9 290 PSU: 550W I am experiencing something that looks to be thermal throttling even though the cpu is not getting particularly hot This is a screenshot from MSI afterburner./\ This is what CPUID HWMonitor is showing me /\ When playing games (BattleField 4) it has a similar outcome in that the temps dip but on BF4 the CPU usage also dips. To stress test the CPU i am using Prime95 which i think is why the usage stays at 100 throughout even though the temps would indicate that it is fluctuating Any help on this would be appreciated.
It suggests the system is likely throttling even when showing signs of thermal throttling. With a base clock of 3.5 and max boost of 4.0, speeds should align with typical usage. Operating below 100% during gaming is expected, as heavy workloads aren't present.
While gaming with the task manager active, I notice the CPU usage stays around 80% (FPS near 120). Occasionally the FPS drops to about 50 and CPU usage falls to around 40% for a short time—this happens even when I haven’t interacted with the game, so it’s not caused by in-game activity. There aren’t noticeable drops in GPU temperatures or usage, which makes me suspect the issue isn’t related to GPU performance. It might be connected to insufficient power delivery to the CPU. I have a four-pin connector in the motherboard, but my PSU uses a different four-pin cable, which I assume would work better if I had a more suitable one.
I believe I'll give it a shot, since that's the only thing I haven't tried yet.
I don’t recommend exceeding 1.6v because it could damage your CPU if you go too high. I use 45nm cores and they’re very reliable, so I can actually run at 2v without harming them. But if you add any stress, I’m worried it might fail, so be mindful of the voltage.
I can handle 2v, but that doesn't mean I recommend it or even plan to push 2v on my e6700. Three times failed attempts and a boot caused by something unexpected mean I'm cautious. So 1.5v seems more sensible. If you have a lower-end board, it should be fine since they rarely reach that level. My G31M S2C only reaches 1.6v, not close to the limits for my 45nm process, so I expect the OP board will likely support a boost of around +200mV or less because it's designed for lower power.
The FX-9590 operates around 1.5 volts, relying on robust liquid cooling to manage its heat. With a temperature of over 60°C and a thermal cap of 70°C, it shows limited margin. Compared to Intel chips, the thermal constraints are tighter, and AMD's 45nm process isn't on the same scale.
Seems like having tough hardware comes with its own set of challenges. I can push two times my capacity if needed, and my processor stays alive even with heavy use. But once you get used to it, you start ignoring the limits and thinking it’s fine to run beyond safe levels. Yes, I managed to boot two times on a cooler with ambient cooling, hitting around 4.6GHz. That’s way more than what most CPUs can handle at safe voltages, especially with modern designs. If I put stress on it, it would definitely fail. Even with a newer chip, I’d probably end up killing it, because I’ve gotten so used to bending the hardware. If someone gave me a Westmere chip, I’d still risk damaging it—32nm is pushing it.