Adjusting OC settings for i9-10900k at 5MHz
Adjusting OC settings for i9-10900k at 5MHz
Even though this was the video I watched the first time I touched overclocking,
View: https://www.youtube.com/watch?v=X26WU900IAQ&t=2814s
The other details I find online don't add much new compared to what I've already seen.
You won't gain much from that video beyond the idea of causing yourself inconvenience. The most useful resource for learning how to customize a 10900k is the Asus ROG forums. What matters is the theory and details you seek, not the specific brand or model. Vcore, LLC, vccio, VCCSA, current high/low specs—these are consistent across any motherboard, BIOS, or vendor, regardless of naming differences.
An engine is an engine, whether called a motor, car part, or anything else. Understand the bios settings, their functions, and how they interact. Focus on the principles; the exact numbers or names may vary, but the overall approach remains the same.
Notably, using CPU-Z instead of aida64 allows it to function at 5ghz with 100% performance and remains stable below 80 degrees... hmmmm
I've also been looking at this and it's quite similar to how I configured everything. The product link is https://siliconlottery.com/products/1090...7018582102. Almost everything matches, except for the CPU Load-line Calibration—where level 3 is replaced with level 4, creating a flat line. This should make the system more stable in different situations.
After reviewing more details, I still have some questions regarding IO and SA. I'm only using 3600mhz RAM, and I've seen that it requires about 1.15v at its maximum (I usually set it to 1.2, which seems recommended for 4000mhz). Some people mention keeping their IO and SA at 1.2 with only 3200mhz RAM. Can I get some feedback on this?
P.S.
By the way, I found that lowering the vcore voltage to 1.29v actually improved stability, with less extreme fluctuations compared to 1.3v.
Aida64 relies on AVX instructions during its testing, which can result in performance reaching up to 130% of the CPU capacity, significantly exceeding typical usage levels. Prime95 small fft serves as a reliable thermal indicator, allowing disabling of AVX features and offering consistent 100% performance, representing the worst-case scenario under regular conditions. A few games utilize minimal AVX, and a rare mod in Skyrim employs AVX2, which is even more intensive than AVX but still considered within normal limits since it doesn’t push the CPU beyond 100%. Yes, Aida64 and similar systems can cause high temperatures during testing, but these outcomes should not be viewed as standard. It’s important to establish clear test definitions. The distinction between stress and temperature tests matters, even if they overlap; using Prime95 small fft with AVX off in a temp test is effective, whereas relying on a single test type for definitive results is unreliable. Temp or stress testing should be combined, but results from any one test should not be assumed conclusive for general use.