Needed assistance during failures urgently!
Needed assistance during failures urgently!
Hi everyone,
I've been experiencing crashes while overclocking, and I'm unsure what might be causing these issues.
My system setup is as follows:
CPU 6600k
RAM Corsair Vengeance LPX 16GB (2x8GB) DDR4 DRAM 3000MHz C15
MB MSI Z170 SLI
GPU GTX1070
PSU 700W Xigmatek
Zalman Z9 Plus
HDD - Samsung 850 EVO 500 GB SSD
EVO 212 Coolermaster
The tower and PSU are several years old, but all other parts were purchased in August. These crashes don't happen during games (though they do occasionally in BF1), usually they occur on the desktop, sometimes during stress tests, but often they just appear without warning.
At first, I overclocked to 4.5ghz using XMP to push the RAM up to its advertised 3000MHz. There were no issues for at least a week. Then instability began. I've experimented with different settings over time, but I don't remember the voltage specs—I've never exceeded 1.35V—and temperatures in Intel Burn Test never surpassed 75 degrees.
For the last six weeks, I've been running at 4.4ghz, RAM at 2600MHz, with a 1.29V core voltage and no problems until recently. Since then, maintaining 4.4ghz has proven difficult, even when keeping RAM at 2133MHz without XMP enabled.
I've also reset the CMOS multiple times after these crashes.
Some people have warned about this motherboard (maybe I chose the cheaper model and overlooked the risks of not having an LLC), and I'm considering replacing it with something better. However, I really value advice from those more experienced in this area.
I don't have access to any other components for testing individually.
Thank you all for your help!
Well that's not acceptable. Crashing during BIOS setup isn't typical. To confirm, you might try loading the latest BIOS again. It also seems reasonable to consider the PSU as a factor. Xigmatek has a reputation for producing less stable PSUs initially, and after some overclocking stress, it's possible they're not delivering consistent power (gaming systems generally need high-quality PSUs for optimal performance, particularly when overclocked): Although I'm aware that some newer models have shown improvement: http://www.jonnyguru.com/modules.php?nam...y&reid=294 Your RAM appears to be in good condition if Memtest ran for an extended period, and the CPU wasn't overly strained.
I should add - when I load defaults in the BIOS, I get no instability at all. The BIOS is the latest version. I realise I may have just massively lost the silicon lottery, but I have also had crashes where I overclock the memory only, which is what led me to believe it might just be a terrible motherboard for overclocking.
I'd really like to know if anyone can explain why I can't reach the overclocks that seemed to work fine and passed the stress tests in the past. Does this mean something is damaged? Can that happen, even if the temperatures haven't reached dangerous numbers?
Thanks again
The journey to a rock-solid stable overclock focuses more on the steps leading to the final settings than on the settings themselves. This detailed manual will explore the various adjustments you need to consider for a reliable outcome: http://www.tweaktown.com/guides/7481/twe...index.html. Often, achieving stability requires more than just voltage, multiplier, and XMP presets. Keep track of the order in which you apply each change. Always test each modification, remember that minor adjustments can make a big difference—especially when fine-tuning at the end.
Hi Geekwad,
I understand your situation. You’ve tried detailed configurations with offsets and adjustments suggested by others for your motherboard, which provided some stability for days or weeks before issues arose. When you began overclocking, starting at around 4ghz and gradually increasing the vcore, you found it took more time and patience than you currently have. Since temperatures stayed manageable and CMOS was reset regularly, could there be another factor affecting your results?
Could you share any details about the stability issues you're experiencing with settings that worked before? It might help to know if you've performed any 24-hour stress tests recently. I noticed some failures during longer periods, which suggests the stability may have been lower than expected.
Geekwad:
Dale7_uk could provide some understanding about why stability issues persist despite previous settings. The main point is that these settings might not have been as stable as initially believed. Have you performed any 24-hour stress tests? I’ve experienced failures lasting from 3 to 13 hours, indicating reduced stability.
Thank you for your response. I haven’t conducted a full 24-hour test myself—my longest was around 4 hours. Now I’m encountering problems with the default settings, including unexpected crashes in BIOS, which hasn’t happened before. This suggests a potential hardware issue rather than a stability concern. I ran Memtest overnight without errors, leaving me to consider the motherboard, CPU, and PSU. Unfortunately, I don’t have spare parts for testing.
If you or others reading this can share thoughts on which component they suspect is causing the problem, it would be greatly appreciated.
Well that's not acceptable. Crashing during BIOS setup isn't typical. To confirm, you might want to try loading the latest BIOS again. It seems reasonable to also consider the PSU as a possibility. Xigmatek has a reputation for producing less stable PSUs initially, and after some stress testing, it's possible they're not delivering consistent power—especially important for gaming systems that need high-quality PSUs for optimal performance, particularly when overclocked: While I'm aware some newer models have shown improvement: http://www.jonnyguru.com/modules.php?nam...y&reid=294 Your RAM setup appears to be a good indicator—consider running Memtest for extended periods. If your CPU isn't being heavily taxed during overclocking attempts, it might point to the board or PSU as the issue. If you could access a PSU, even one that's underpowered for your GPU: http://www.realhardtechx.com/index_archivos/Page362.htm You could eliminate the problem by removing the dedicated GPU and testing with the integrated one.