Analyzing Stress Test OC Configurations/Prime95 Challenges
Analyzing Stress Test OC Configurations/Prime95 Challenges
Hi everyone, how are you approaching stress and thermal testing for a CPU overclock today? I just completed a new WC loop and want to verify its success. No need to adjust the CPU cooler or TIM settings—just ran Prime95 on the Small FFT profile. Is this still a reliable measure of both loop performance and system stability? On the overclocking aspect, I’ve heard Prime95 can adjust its voltage to match your OC preset. Since I disabled voltage changes in the BIOS, will that affect the results? Am I actually seeing the CPU stress-tested at the voltages I set, or at the ones Prime95 requires?
Check the Asus Real Bench leaderboard at https://rog.asus.com/rog-pro/realbench-v2-leaderboard. If it meets the 8-hour requirement and only AIDA64 default and FPU run, everything should be okay.
Your setup is performing well at 4.8Ghz with a stable 1.275v core voltage. To potentially increase performance, consider adjusting other parameters such as clock speed, power limits, or even exploring different chip models. The decision to exceed 1.3v depends on your specific needs and the chip’s capabilities.
I don't understand why the system wouldn't support higher voltage. BIOS typically allows a wide range, even though it risks damaging the chip. Either way. 4.8 GHz on Haswell is quite fast, so if it runs steadily, you should be okay. I used a 4770k at 4.4 volts, but I don't remember the exact voltage. I recall a 1.25 and that was on liquid metal with a custom loop. It probably had more power too, but I'm not sure. Shortly after, I got a 6700k at 4.6 GHz, likely around 1.25 as well. The main point is that a 4770k at 4.8 isn't significantly slower than a 7700k at 4.8. That's still a strong CPU, and pushing to 4.9 or 5 wouldn't make much difference. I'd check if you can achieve stability at lower voltages, otherwise it's fine.
Sorry, I didn't mention my CPU model before. I own a 4790k. After some research, I found it set to 1.7v and it wouldn't move closer than 0.4v. Eventually, I stabilized it at 4.9 with a 1.325vcore. However, one core overheated during Prime95 tests, so I lowered it to 4.8 and got the lowest stable voltage at 1.265. For RAM, I kept it at the original spec of 1.65v to maintain stability. If you adjust VCCin manually, like setting it to 1.7v, could that affect performance or cause problems?
Man, I don't recall DDR3 RAM configurations at all. If you try 1.65v on DDR4, you're probably getting a terrible result. Haha. I really don't remember... Probably some detailed discussions on the extremeOC forum about DDR3 overclocking.