Voltage response of the overdrive during precision boost with R5600X and Asus BIOS
Voltage response of the overdrive during precision boost with R5600X and Asus BIOS
PBO Scaler measures how much aggressive the single core frequency is compared to all cores. Essentially, it shows how much the single core speed drops versus the multi-core speeds. The maximum CPU boost clock override raises the upper limit for allowed frequency, though this value differs between Zen 2 and Zen 3. For Zen 2, PBO ensures safe voltage levels even under stress. With low workloads, voltages might reach around 1.47V, but higher loads cap at about 1.35V. You can experiment with tools like Prime95 smallFFT and adjust thread counts to see effects. It seems AMD keeps this behavior stable on Zen 3 using 7nm+ processes, which handle high voltages better. For similar reasons, Ryzen boosts come mainly from effective cooling, not just clock speed. If you prefer more conservative voltage control, consider using offset mode with a negative value (e.g., -0.05V) to lower the voltage by 1.45V to 1.4V. Always double-check performance after making changes. This is Intel's approach.
Perhaps my caution or lack of understanding about CPU overclocking is the issue. If under "low thread count workloads (where the CPU stays cooler) you might notice up to 1.47V," isn't it risky for these specific cores? Does it really depend on both voltage and temperature for degradation to happen? Based on my three years of experience with daily overclocked Ryzen, I saw those numbers and they concerned me. I read about the "observer effect" where monitoring tools show higher voltages than actual values, especially during idle. I thought my situation didn't match because I was under load, but now I see it clearly. I'll test with Prime to confirm. Umm... Wut? Maybe this was meant as a joke, but I really didn’t get it. Sorry.
Degradation happens under high voltage and heat, which isn't inherently risky. Core OC stays at 1.4V or higher, but that's hazardous. Intel's turbo response is quite simplistic—it just scales cores based on usage thresholds. It lacks flexibility and is designed for the most extreme CPU configurations. For clarity, the initial PBO version (on Zen+) performs even worse, often needing excessive voltage and relying heavily on offset mode (like pushing 1.55V on a 90°C+ 2700X setup).
Negative offset functioned well, especially in single treaded cases where the sustained voltage was around ~1.37V. From a performance standpoint, results were consistent across short tests I conducted. Appreciate the guidance. I had hoped every maker would adopt their turboclock approach this way. It's a cautious strategy: inform customers about clock boosts beyond what they expect, even if the CPU example is subpar. They’ll feel confident knowing they received value, and marketing teams will likely celebrate this success.