Yes, C-states play a significant role in the system.
Yes, C-states play a significant role in the system.
I understand this might be a beginner question, but I noticed many mixed messages online, so I thought I’d share here. A few days back my motherboard (ASRock B650M Pro RS) began making a loud buzzing sound, stronger than other components. After some testing, I realized it was related to the VRM coil whine. I searched for fixes and saw one suggestion: disabling C-states in the BIOS. That worked—right away the noise disappeared. From what I know, C-states let the CPU partially shut down when not actively using it, which helps save power. Using HWiNFO, I recorded the varying power usage. My findings indicate the biggest impact happens when the system is almost idle (18-20W with C-states on, 25-27W without), and even a simple mouse movement can push it up to 30-33W (or more if doing other tasks). I also read that C-states keep the CPU locked at a high clock speed, but for my setup—CPU runs at 4941 MHz most of the time (with occasional drops to 4283 MHz or higher, and rare spikes to 5140 MHz)—it stays there regardless of whether C-states are enabled. I have a few questions: - Do C-states influence power consumption when the system isn’t idle? - Does it really change clock speeds? If so, why don’t I notice that on my machine? - How significant is the power difference? I’m not much concerned about it yet, but I want to know if it could matter financially if I leave the PC running all day.
C-states aren't crucial in real-world usage with the PC. Ryzen might show an observer effect with monitoring tools. Generally, the CPU could confuse the software for a background task and increase speed while you're watching it. Using Ryzen Master should let you notice downclocking when C-states are active. Leaving the CPU idle when not in use usually makes a difference if C-states are turned off.
I'll check out Ryzen Master tomorrow, appreciate the suggestion.
Twelve years prior, a major issue existed with Intel's Bay Trail (budget laptop chips) line because certain operations could unexpectedly freeze and halt any non-Windows OS. Disabling C-states became essential, but it also caused significant power consumption increases due to other architectural problems. This was the final time C-states played a meaningful role.
I understand, thanks for your feedback. My tests with the Ryzen utility gave me some intriguing findings. The power variation between on/off states matches what HWiNFO shows (5-7W), but the actual numbers look different—HWiNFO added roughly 15W to the readings. I might be confusing how these programs track clock speed. Both HWiNFO and Task Manager consistently display around 4.91 GHz, while Ryzen Master fluctuates between peaks of about 100 MHz when idle and 1000-2000 MHz during video playback. Given the CPU’s base speed is 3.8 GHz, the reported peak speeds seem low compared to what I expected. It seems HWiNFO might enhance performance on its own, but the data from Ryzen Master is significantly lower. Also, it often marks cores as idle even when C-states are off. Likely, I’m misinterpreting the information.