i5 6600k overclocking - bsod
i5 6600k overclocking - bsod
Hi!
I recently got a new rig about six months ago using an i5 6600k, an ASUS Z170 Pro gaming PC, and a Hyper X Savage 2x with 8GB RAM at 2400MHz CL 12 in XMP mode. During stress tests with AIDA, it was running at around 4.4GHz with manual settings at 1.25v. Everything worked fine, even for demanding CPU-heavy games like Battlefield 4. Recently, after about three hours of gameplay, I got BSOD in Mankind Development after 15 hours of playtime. I increased the Vcore to 1.265, but it still didn’t happen. However, when playing Overwatch for about 15 hours straight, I encountered BSODs or freezes accompanied by buzzing sounds, which forced me to restart. This only occurred when minimizing the game and browsing the web—never during actual gameplay in this title. Today I decided to set default BIOS settings and didn’t see any BSOD when minimizing (though my turbo was at 3.6GHz instead of the expected 3.9GHz).
I also play games like Witcher 3 without any issues. Can someone help me understand why this happens? I don’t want to increase voltages further, as my max core temperature reported by SpeedFan is 67°C at 1.28V, but it only reaches 52°C when running at default.
Memtest is designed exclusively for memory or RAM checks. Avoid running it inside Windows since it loads into memory during operation, which might indicate a software problem if it reports an issue without providing a clear solution. The BIOS on your Asus motherboard includes its own Memtest feature that can be executed independently of any operating system, offering a more reliable test. If no errors appear, proceed to alternative troubleshooting steps. Consider re-seating components and power cables, including RAM, the video card, and any physical connections. Also, reapply thermal paste to your CPU cooler. If temperature spikes are responsible for the BSOD, inspecting the cooler and ensuring proper airflow should follow. What CPU cooler did you install?
If you revert to stock configurations, the chip might struggle with the overclock. It’s better to return to stock and check if the same issues arise. If they persist, try increasing the clock speed slightly and test again.
Regarding cooling, consider whether your CPU uses water or air cooling. A low-quality cooler could be a problem. Ensure proper airflow within your case, as this can also affect performance. Additional information on these aspects may be useful.
as i mentioned, bsod occurs exclusively in this specific situation
Getting more volts is essentially the right approach, whether you're aware of it or not. If things get too warm, you'll need improved cooling. The 4-core turbo model is only 3.6, which explains why you noticed that. At stock, the 3.9 single-core turbo is available.
When i bought it, it turbo to 3.9 ghz, now even when watching 4k video on yt it turbos max to 3.7ghz. Regarding voltages, I adjusted them to 1.285v. The issue is that different programs display varying values. cpu-z always shows 1.280, hdmonitor doesn't show anything, and only vid and AI3 show 1.280 and 1.296 when under heavy load. It's unclear why it isn't showing 1.285 as I manually set it in bios. Also, could the RAM XMP profile be affecting this? If the CPU was unstable, it would crash in every CPU-heavy game, not just when the game is paused and resource usage is low? I've run Windows memory diagnostic before, but the results were inconclusive. I also tried memtest, but it only reported errors on my pendrive without any action.
XMP configures your CAS value in memory according to the profile supported by your RAM. If you're uncertain about manually adjusting the CAS setting, simply rely on the XMP options and it should work adequately. Make sure to update your BIOS and all relevant drivers if you haven't already.
Gazi :
i have newest bios. I think it might be ram issue. i have hyper x savage 2x 8gb 2400 mhz cl 12. no hyperx are listed on the ram supported list. when i set those to auto(downclocked) there is no bsod when alt tab so far... .
HyperX isnt bad. Should have a decent warranty too. This is why I was saying run memtest though. Running that a few times should confirm if this is the right point of failure or not. Run it with XMP off and then on as well to see if that affects it.