Static overclock vs PBO
Static overclock vs PBO
Static overclock performs better for me compared to PBO, though PBO is still suggested over static OC. My 5600X is fixed at 4.8GHz across all cores with SMT turned off; I used to run 4.9 but the power demands were too high. With a better board, I could reach 4.9 or even 5GHz. It outperforms PBO in stability, especially in gaming where frame rates are more consistent. Lower frequencies feel smoother, temperatures drop, and overall desktop performance is faster. Input lag is more responsive, and everything runs snappier.
It performs reliably within 5% in multi-core setups and tends to run a bit quicker on single cores. The 5900X remains stable around 4.55GHz with all-core optimization, though both CCDs show similar performance. When PBO is enabled and certain limits are applied (like 188, 177, or 222), it reaches 4.6GHz under load—just a minor clock stretch. In Y-Cruncher 1B, it lags by about 200ms compared to a fully stable all-core version. Compared to Cinebench R23, it edges out by roughly 200 points, though the exact figures aren’t critical. Turning off SMT could push it another 100MHz, but the all-core version still shines. Running at 4.8GHz on two cores is achievable with mild stretching, giving a solid edge over an all-core build. If you can maintain high voltage (the clock drops above 1.35V without SMT), and run at that speed, PBO might not be necessary—though it can boost single-core performance to 5.1–5.2GHz on a few cores, offering a noticeable gaming advantage. Overclocking is mostly for testing; beyond that, static boosts are mainly for records.
It's more complex than it seems; the device remains stable. AMD adjusts the frequency every millisecond based on temperature and voltage to the chip. If your regulator isn't precise, clock stretching may increase slightly. The performance is still present, which is the main concern. This is per chip—my sample from March 2021 is early and not very reliable (4.75GHz max for Time Spy Extreme, only on a cold start). You might try lowering the voltage to test stability, but results won't be promising.