High VRM temps?
High VRM temps?
Hello. I'm trying to understand what is considered "high VRM temperatures" for MOSFETs and the VRM controller. My board is a Gigabyte AORUS Gaming K3, and when I set the LLC to "High", the next level Turbo didn't work, but the VRMs still get very hot—reaching 88°C in LinX monitoring. I stopped using it. My BIOS settings are:
vCore : 1.260v
LLC : High
VCCIO : 0.950
VCCSA : 1.050
XMP Profile (3000MHz, 1.35v) on a Corsair pair of 2x8GB RAMs.
For tracking VRM temperatures, I'm using the latest stable HWINFO64. My CPU is cooled by Scythe Ninja 4 with paste MX4 in Fractal Design Define R5 Case, featuring two intake and one exhaust fans running at full speed. I haven't adjusted any temperature settings, just letting it run at maximum performance. Despite this, the VRMs are still getting such high temperatures.
I still have the option to return the board to the shop and replace it with something else in that price range. Could it be that the VRM cooler isn't properly secured—maybe it's a plastic lock or similar, not just screws? What do you think about monitoring VRM temperatures 24/7? Is that considered safe?
According to common references, a standard maximum temperature for power MOSFETs in VRMs is around 125°C. Anything below 100°C provides a significant safety buffer. It seems you checked the VRM temperature during a CPU stress test or benchmark, which is a more realistic scenario than typical operation.
These controlled temperatures were subjected to a rigorous stress test to verify the stability of my CPU overclock. It was operating with LinX and 14GB of RAM allocated. While I don’t require overly aggressive overclocks, I aim for a mid-level overclock reaching around 4.7 to 4.8 GHz. Perhaps adjusting the AVX negative offset could provide extra assurance without risking any issues. Certain games and applications utilize AVX (as known from CS:GO, GRID, and possibly others). Occasionally, I render videos and spend time with music daily.
Regarding the VRMs, a maximum of 125°C is acceptable. I’m concerned about potential lifespan effects or damage, especially if temperatures exceed this. Ideally, keeping them below 100°C would be best, even if it’s not immediately harmful. The ambient temperature is currently around 22°C, but could rise in summer. It’s important to maintain some buffer for temperature fluctuations during these periods.
Is this approach safe? Would a sustained temperature of 90°C continuously pose any risk to the lifespan of the MOSFETs or VRMs over time? I’m hoping to use this board for at least two years, or longer, depending on future Intel CPU releases.
It seems the FETs can handle junction temperatures up to 150°C. Therefore, maintaining a VRM temperature of 90°C during maximum load is perfectly acceptable for continuous operation.
Thank you for your response and the details you provided. I also found datasheets for MOSFETs, but there’s nothing about used VRMs—any kind of guidance on their survivability. Likely I’ll aim to keep temperatures below 90°C under heavy stress. Under normal operation it should be fine, even with CPU-intensive tasks, since high usage probably won’t require AVX instructions and wouldn’t push the system into extreme conditions. It’s a bit confusing because some sources mention staying under 80°C, but 90°C should be acceptable if not continuous. Thanks again for clarifying.
The VRM's temperature reading would likely reflect the MOSFET temperature. The controller probably wouldn't get much hot, and the inductors should remain stable in temperature. The capacitors might warm slightly, but the MOSFETs are the main focus of monitoring.