I5 8600K Stable 4.8Ghz Overclocking Configurations For Gigabyte MB
I5 8600K Stable 4.8Ghz Overclocking Configurations For Gigabyte MB
Do you have the memory settings set for XMP or other custom configurations? I already described what to do if it fails. It's in XMP Profile 1 at 3000Mhz, and I just changed the voltage. I'm asking how I'll know if it didn't pass the test—will the PC restart afterward?
Typically when running Realbench, you'll notice the CPU usage drops to zero. If it does, it will indicate that. You can adjust the time and memory limits. I recommend setting half of your RAM in the memory configuration. Remove the XMP profile from the memory and revert to its default settings. Restart CMOS and reapply your CPU configuration. It's best to keep memory at its standard, basic level during testing before overclocking the CPU. Once you confirm the overclock is stable, you can proceed with memory adjustments. Otherwise, memory problems might lead to an incorrect stress test result. Always prioritize CPU overclock first, then consider XMP or custom settings.
Darkbreeze :
When running Realbench, you typically notice CPU usage dropping to zero. If it continues, the system will indicate this. You can adjust the time and memory settings. I recommend setting half of your RAM in the memory configuration.
I suggest removing the XMP profile from the memory and restoring it to its default values. Restart CMOS and then reapply your CPU settings. For testing overclock stability, keep the memory at its standard basic configuration. Once you confirm the CPU overclock is stable, you can proceed with memory adjustments.
Otherwise, memory problems might lead to an incorrect stress test failure. Always try CPU overclock first, then consider XMP or custom settings. Not the reverse.
When I turned off the XMP profile, I saw a blue screen appear after the motherboard logo loaded. After increasing the DRAM voltage again and applying XMP, it booted up successfully? This seems normal.
P.S. 1 hour Real benchmark completed at 79°C max temperature. LLC set to standard.
Temperatures in Realbench aren't important. You're not checking temperatures, you're searching for errors. Any result under 8 hours is irrelevant too—whether it's 1 or 5 hours, it doesn't matter. If you fail the 8-hour test, the system isn't stable. Just because it doesn't trigger an error in a short time doesn't guarantee stability. If you can't run it for the full 8 hours, it's best not to attempt it until you're ready. Also, this isn't typical behavior. I'd recommend starting fresh. Remove the CMOS battery from the motherboard while powered off and unplugged. Wait five minutes, then reinsert the battery, restore power, adjust CPU settings in BIOS for overclocking, but leave memory settings untouched. Return to Windows and run Realbench for 8 hours. If it succeeds, you can proceed with setting your XMP profile and required voltage. After that, perform additional memory tests using Memtest86 for four passes. If successful, continue with Prime95 version 26.6 Blend or custom settings (512k min FFT, 75% RAM usage). I can walk you through the exact steps if you reach that point.
Darkbreeze:
The numbers in Realbench aren't important. You're not checking temperatures, you're searching for issues. A test shorter than eight hours is irrelevant too—either one hour or five hours counts the same. If you fail the eight-hour check, it means the system isn't stable. Just because it doesn't throw an error within an hour doesn't matter much. If you can't run it for eight hours, it's probably best not to try it at all until you're ready.
Also, this isn't typical behavior. I'd reset everything. Remove the CMOS battery from the motherboard while power is off and unplugged. Wait five minutes, then reinsert the battery, restore power, enter BIOS, and adjust your CPU settings for overclocking—don't change memory settings. Return to Windows and run Realbench for eight hours. If it succeeds, you can then configure your XMP profile and the required voltage.
After that, perform further memory tests with Memtest86 in four passes. If successful, proceed with Prime95 version 26.6 (Blend or Custom settings: 512k min FFT, 75% RAM usage).
I can walk you through the exact steps for testing RAM in Prime if you reach that stage.
I tried it for four hours and it passed, but I still need to extend it to eight hours. If it fails then I'll try again. I really appreciate your guidance on the RAM—thank you for all the help! No one else in the forums helped me as much as you do.
Darkbreeze: I also just wrote this from the memory perspective. It could be useful too. You can find more info here: Hello, I didn't reply because I wasn't at home to test it, but I did test it today and passed the 8 Hours RealBenchmark Stress Test for Stability. https://imgur.com/a/XTv0NHK Should I increase the voltage for better stability or should I keep it as is? Thanks a lot—I really appreciate your help so I can make this thread resolved!
If it is stable, I see no reason to increase the voltage. Maybe increase by .005v if it makes you feel better.