I'm trying to maximize the lifespan of my Lenovo Legion 6, which has a Ryzen 7 5800H and an RTX 3060. I use an external cooling pad and a custom fan profile capped around 2500–2700 RPM. Turbo Boost is disabled, keeping the CPU near 30 W under heavy load, while the GPU is undervolted to 0.750 V at 1550 MHz and draws roughly 70 W. During demanding games, both core and hotspot temperatures remain below 60°C. The laptop doesn't expose dedicated VRM temperature readings, though. Could the CPU or GPU power-delivery components still run hot or experience thermal stress despite such low core temperatures? Is limiting the internal fans to around 2500–2700 RPM safe for the VRMs, or should I allow higher fan speeds near the system's thermal-throttling point?
3 Answers
An external cooling pad doesn’t automatically cool every internal component. Its effect depends on where the laptop’s intake vents and VRM heatsinks are located, and in some designs an external fan can even alter the intended airflow pattern. If the pad is clearly lowering CPU, GPU, and SSD temperatures, it’s probably helping overall, but that still doesn’t provide a direct VRM measurement. With your reduced power limits and sub-60°C core temperatures, there’s no obvious sign of dangerous VRM stress; just keep an eye out for throttling, instability, or unexplained shutdowns.
Low CPU and GPU temperatures make VRM overheating much less likely, especially because you’ve reduced both CPU and GPU power draw. However, core temperatures don’t directly prove the VRMs are cool—their temperature depends on the laptop’s heatsink and airflow design. If the VRMs sit beneath thermal pads connected to the cooling assembly, they’re probably receiving some heat-spreading and cooling benefit. Firmware may also monitor or protect power components even if the readings aren’t exposed to the operating system. I wouldn’t add aftermarket heatsinks unless there’s an actual problem, since clearance and pad contact can be easy to get wrong.
Your GPU undervolt should generally reduce VRM stress because the GPU is using less power, although the exact temperature reduction depends on current draw and the VRM design. The main thing is stability: test the undervolt thoroughly, since an unstable setting can cause crashes even when temperatures look excellent. For the fans, a fixed 2500–2700 RPM limit may be fine at these power levels, but leaving a higher-speed range available near the laptop’s thermal limit gives the system more protection during unusual workloads.
I checked the internal layout and the VRMs appear to sit beneath the pink rectangular thermal pads, so they seem to be part of the intended cooling path. That makes me less concerned about them running completely uncooled.