Initial attempt at boosting performance Let’s give it a try!
Initial attempt at boosting performance Let’s give it a try!
I've experimented with OC settings and am still unsure about the results. My build includes an Asus X570 TUF Gaming Plus with a Ryzen 5600X, Corsair Vengeance 3600C18 (2x16GB). CPU temperatures stayed between 75-76°C. I managed to push my core clock up to 4.5GHz with little adjustment, which is a bit higher than the 4.4 reported elsewhere. I haven't adjusted the single-core boost yet. My RAM reached 3733 MT/s at 1.41v, but lowering it caused instability—sometimes showing 18 instead of 17. I'm currently at work and want to verify the voltage, aiming for a slight increase to around 4.6GHz. It might be worth it if it helps.
For AMD CPUs, PBO is a convenient option that delivers around 4.5-4.6GHz without the need for manual overclocking. It avoids dropping single-core speeds compared to full core tuning. Turn off all other overclocking configurations, activate PBO, and you're good to go. If you're comfortable with fine-tuning, you can experiment with underclocking and the curve optimizer for even better results—especially if Gear Down Mode is enabled. For those just starting with memory overclocking, keep Gear Down Mode on; it simplifies stabilization without significantly affecting performance.
Focus on CPU performance by enabling PBO and applying the curve optimizer to lower the voltage. This allows the processor to operate at higher single-core boost speeds and static OC levels. Keep in mind that RAM specifications vary per chip—use a tool like Thaoon Burner to identify your motherboard model. Generally, aim for around 1.5V max unless cooling is insufficient. Actual safe voltages depend on the specific chip; research accordingly. Chips such as Micron, Samsung, and Hynix can handle 1.65V without problems if proper cooling is in place. However, some models like the Samsung C Die may drop to around 1.35V, though this seems questionable. For chips that ignore voltage stability, push them to their limits, but expect failure beyond 2.05V. Avoid exceeding 1.7V just to prevent degradation. If your RAM supports scaling, it’s less critical; otherwise, stick below 1.5V for best results.
Yeah, not great not being able to reach 3800. Should we boost voltage stability to fix timing? Or should I slow down to make it work? Or is it just about settling for what you have and tweaking until things go wrong again? Sounds like you're still learning the basics.
maybe some mobo makers... i don't know, my ram wouldn't start at 17 but works at 16 for instance... or 1.35v... no, it should be 1.36v even with XMP 1.35v fine... basically this stuff usually depends on the motherboard.
Looks good, I'll look into it once I'm back. Fingers crossed it's not just dog-crap.
C18 at 3600 is likely from SK hynix. Generally tighter clock speeds for high-frequency designs won't work well with voltage scaling. SOC 1.08V supports XMP, max 1.20V, but extra voltage usually generates more heat than it saves in frequency. Finding the right settings takes a lot of trial and error. Don’t lose hope next week while you’re still adjusting, and likely the following week too. Good luck!
I have confirmed this after investing several weeks to make my 1520mhz ddr2 stable initially, then improving performance by adjusting secondary and tertiary settings a couple of times. I managed to achieve solid speeds around 1520 with a 7-9-6-9 pattern at 2.38v. I haven’t run a benchmark test yet, but I’ll look for one soon. It should work well daily thanks to the low voltage, though I’d like RAM that supports up to 2.95v since lower voltage means less performance gain. Currently I’m planning to add an e8600 to prevent it from dropping to only 4.25ghz, as even a small drop in voltage significantly reduces efficiency. The current cooling setup is fine for now, but tuning it further is necessary because the unit gets too hot and crashes after around 2.2v.