CPU degredation?
CPU degredation?
Hello everyone,
I have a question. I own a Ryzen 5 3600 with an AIO water cooler for months now and have been using PBO or auto OC. Recently, I tried manual OC with a multiplier of 42, but it wasn’t stable until I increased the voltage to 1.42. During gaming, temperatures stayed between 50-55°C. It’s concerning because most people on YouTube got it working in under 1.35V, and someone mentioned it could damage your CPU.
What should I actually do? Should I lower the OC or go back to the default settings? Any advice would be appreciated.
Regards
It's nearly impossible to determine if the device is degraded due to the extensive activity involved. All components—memory, GPU, CPU—are being heavily utilized. Additionally, there appears to be poor coding by the game developer. The recommended approach is to perform an extreme CPU stability test using tools like Prime95, focusing solely on the CPU. Run this for 20 minutes to an hour with a single FFT size to identify the lowest stable voltage at your overclock. After that, repeat the test once more when you suspect degradation, again for 20 minutes to an hour. If it fails to maintain stability, further investigation is needed.
I assessed stability by considering the differences between systems. Since each CPU and motherboard varies, direct comparisons aren't useful unless you replicate their exact configuration. A solid conclusion required a thorough baseline test to confirm if your CPU could no longer sustain performance at the same voltage.
In Buildzoid's situation, he was using a fixed voltage with a high clock speed and heavy AVX processing, which raises concerns. For your setup, if you were running PBO with automatic adjustments, it would have allowed voltage changes based on temperature. Gaming demands minimal processing power, even when timing is critical.
At around 1.4v with alt+tab in the game and YouTube open, my system crashed. Below that threshold I experience repeated blue screens under stress, similar to the cinabench test. After switching back to PBO with a voltage offset of -0.0750, running Call of Duty in the background on the Ryzen balanced power plan resolved the issue.
It's nearly impossible to determine if the device is degraded due to the extensive activity happening at once. All components—memory, GPU, CPU—are being heavily utilized. Additionally, the coding by the game developer appears to be subpar.
The best approach would be to conduct an extreme CPU stability test using software like Prime95, focusing solely on the CPU. Run it for about 20 minutes to an hour with a single FFT size to identify the lowest stable voltage at your overclock setting.
Once you suspect degradation, repeat the same test for another 20 minutes to an hour. If the device cannot maintain that same minimum stable voltage, then it's likely degraded.
Please note I'm not claiming you haven't experienced degradation, just that certainty is still lacking. Running a fixed clock and voltage above 1.2V will almost certainly accelerate wear.
On PBO voltage offset - 0.750, the prime did not crash; maximum voltage seen was 1.3, most often around 1.27v. Clock speed was 3873mhz, with temperatures reaching 72°C. It appears the CPU operates at 1.4v when idle and lowers voltage under load. What advice would you like?
Are you certain the negative offset is -0.750? Might it actually be -0.0750? If that's the case, it's a substantial negative offset and likely leading to performance issues even if you don't notice them in clocks. I'd check clocks and voltage in AUTO mode to see how it behaves. At the same time, ensure the following are enabled in BIOS: CoolnQuiet, Global C States, processor CPPC, processor CPPC preferred cores. Regarding PBO, the processor's boost algorithm continues to monitor voltage and frequency against its FIT tables as temperature rises to prevent degradation. However, it functions accurately only when offsets aren't applied, which is why keeping those settings in AUTO offers a clearer picture of its operation. Lastly: each processor's FIT values are distinct, meaning its performance with varying voltage and frequency across temperatures is also unique. Because all these factors depend on temperature, understanding the ambient temperature is essential since it fluctuates with seasons and affects safe voltage and frequency ranges. This aspect is particularly crucial for air cooling, as liquid cooling offers a larger safety margin.
Thank you for the reply.
The results were consistent with a stable performance at low voltage, but instability appears when the clock speed increases to 4.2ghz while maintaining 1.3-1.35v. The drop in stability suggests possible limitations under higher frequencies.
Is this typical behavior?
Ryzen appears to prefer higher voltage settings. My 2700x tends to struggle when pushed above 1.35 under minor overclocks. Another point is the motherboard's power delivery problems; I previously used a P5B Deluxe board, overclocked my Q9400 to 3.6, and it worked well for a while before occasional blue screens appeared. I downclocked it temporarily, which stabilized it, but issues returned until I moved it to another system where everything functioned normally.
It's common knowledge that you can significantly undervolt Ryzen 3000 processors without notice, as this affects clock speed or frequency. I recall a YouTube reviewer who performed this at launch and had to adjust his review after realizing the impact. It seems similar might be happening in your situation.