Performance peaks at lower workloads, allowing better heat management.
Performance peaks at lower workloads, allowing better heat management.
Hey, I've just upgraded from a R5 3600 to a R9 5950X recently, added more RAM and a CPU cooler (switched from DarkRock Slim to Deepcool AK620). I noticed something odd—the CPU always felt hotter during games or idle than expected. I thought my expectations were wrong, but after testing, it seems that tasks using just 2-4 cores at around 60% usage make the CPU run much hotter, causing fans to speed up. When I ran a full load with Blender and all 16 cores at 100%, the temps dropped back down. TL;DR: 2-4 cores at 60% = 75-80°C, 16 cores at 100% = 55-60°C. Anyone have a clue? Is it because the temperature sensor is closer to the cores used in lighter loads? Are there different core types, like those found in GPUs for 3D, copy, or video encoding? Could it really help if you could?
Verstanden, ich übersetze gerne für Sie. Wenn etwas nicht klar ist, sagen Sie einfach, ich passe mich Ihrer Sprache an.
It seems you're checking on your CPU performance. The faster the CPU and fewer cores it uses, the quicker it can handle tasks. Your Ryzen CPUs handle high temperatures around 90°C. What is your fan curve? 3600 and 5950x. Which power setting are you currently using?
The fan curve remains steady around 20% until temperatures reach 50 or 60°C, after which it increases gradually in a proportional manner until it hits 90°C. I’ll verify later that my power settings are balanced, matching the picture, though I’m not aware of the AMD Ryzen Balanced profile. I also don’t have the Ryzen Balanced setting installed.
I didn't perform any updates before or after a CPU swap. The task only involves discussing the swap itself.
NZXT CAM is already installed, making it hard to identify a problem. Completely remove it, restart the system, and avoid installing it again. Even if it doesn’t affect you now, it remains a threat—potential malware, resource drain, and hidden spyware.
Cam may affect how fans operate correctly. It’s recognized to lead to overheating because it’s less efficient than a brick. It makes sense that if you compare fan speeds without the cam of other fans, interference could occur and cause fans to run much slower.