3600 OVERCLOCK LIMIT? "Is there a clock speed i should stop at regardless of the test still passing?"
3600 OVERCLOCK LIMIT? "Is there a clock speed i should stop at regardless of the test still passing?"
Hello, your setup looks solid so far. At 4.6Ghz with stable temps under 65°C, you're in a good range. If your tests keep passing, consider capping at around 4.8Ghz or 5.0Ghz for better stability. Keep an eye on temperatures and voltage consistency. Let me know if you need more guidance!
Most overclocking experiments don't demand extreme precision regarding faults. Instead of doing it manually on your 3600, rely on the Ryzen Clock Tuner (CTR) with a 1usmus adjustment. Let its self-adjusting boosts take effect, then apply OCCT to evaluate the stability. If you encounter AVX issues, I wouldn't fully rely on it; reduce the clock speeds until things settle. A single error per hour might be fine, but instability will become obvious when errors rise rapidly. You can experiment with voltage settings in CTR until you find a steady point. Even if it passes SSE checks, if overall performance is erratic, it's not worth pursuing. OCCT also covers memory, GPU, and GPU memory tests, which proved useful during EVGA Precision X1 testing (I didn't OC my memory since I'm content with my DOCP settings). Using CTR lets you try a lower voltage setting to save power while still gaining a modest improvement. All in all, once you've settled on a stable configuration, run OCCT for several hours to confirm reliability. Keep monitoring temperatures closely—my fans are managed via Argus Monitor, and even minor BIOS adjustments didn't improve the balance between temperature and noise. While it's entertaining to tweak components, overclocking the 3600 offers limited gains compared to top-tier CPUs from AMD and Intel. If you're curious about my results, I have a Silver Sample; my typical settings are VID 1286, CCX1 4300, CCX2 4325. My baseline underclock is VID 1175, CCX1 4150, CCX2 4175. Those were the most successful configurations after countless adjustments. Overall, it was enjoyable!
The issue with version 1.375 might stem from the constant behavior, possibly due to AMD's max save voltage of 1.5 and the chip's actual limit around 1V. It sounds like the hardware is already capped there, which could explain why it feels redundant. Tech advice seems to agree on this point.