6600k oc with cm hyper 212x isn't stable.
6600k oc with cm hyper 212x isn't stable.
My system with a 6600k processor, 212x hyper clock, and 4.6ghz voltages at 1.375 ran AIDA64 smoothly with stable temperatures for 12 continuous hours (peak temps reached 62°C, while average across all cores was around 55°C). When launching GTA V, it performed well at 60fps for about half an hour to an hour before crashing. I’m looking for a solution to resolve this issue. Thank you!
AIDA64 is a poor stress tester. Try the one at http://www.techspot.com/downloads/4965-i...ntest.html for better results. Use http://www.cpuid.com/softwares/hwmonitor.html to monitor temperature and vcore in Windows. You might want to raise your Vcore by 0.010V and test again. Keep repeating this process if you experience crashes. Make sure to start with manual voltage until you reach a stable setting. Also, 1.375 volts should be sufficient for 4.6; checking your Vcore in Windows is important since it may drop during load or gaming.
AIDA64 is a poor stress tester. Try the one at http://www.techspot.com/downloads/4965-i...ntest.html for better results. Use http://www.cpuid.com/softwares/hwmonitor.html to monitor temperature and vcore in Windows. You might want to raise your Vcore by 0.010V and test again. Keep repeating this process if you experience crashes. Make sure to start with manual voltage until you reach a stable setting. Also, 1.375 volts should be sufficient for 4.6; checking your Vcore in Windows is important since it may drop during load or gaming.
Makentox:
AIDA64 is a poor stress tester. Try the alternatives at http://www.techspot.com/downloads/4965-i...ntest.html for better results. For monitoring temperature and vcore in Windows, use http://www.cpuid.com/softwares/hwmonitor.html. You might want to raise your Vcore by 0.010V and test again. Keep checking your manual voltage first until stability is reached. Also, 1.375 should be sufficient for 4.6; it's important to verify your Vcore voltage in Windows since it may drop during load or gaming. Thanks for the advice, I'll try them tomorrow.
Various stress tests function uniquely and might miss certain situations you'll face in real-world programs such as games. It's typical to successfully complete one or two tests without issues for 30 minutes to several hours of testing, after which you start another test and crash within 5 minutes or less because the test uncovered an instability that others overlooked. Another versatile stress test for stability is rog realbench. It evaluates all components of the PC—CPU, RAM, GPU, and memory subsystem—through a range of tests that more closely mimic actual usage patterns.
Different stress tests function uniquely and might miss certain scenarios found in real-world programs such as games. It's typical to successfully pass one or two tests without issues for 30 minutes to several hours of testing, then immediately move to another test and crash within 5 minutes due to an instability discovered during the process. Another versatile option for stability is rog realbench. It evaluates all system components—CPU, RAM, GPU, and memory subsystem—through a range of tests that more closely mimic actual usage patterns.
What suggestions do you have for the safe maximum voltage? Keep it below 1.4v and under 80°C continuously. Monitor your temperature closely during a stress test—it will rise significantly during an intel burn test.
Plus one for Makentox's advice. A modern CPU should limit performance when overheating but acts more as a protective measure. It's best to monitor temperatures during stress tests, particularly when overclocking. Avoid setting thresholds like ibt or occt (linpack) and then shutting down or watching TV.
I'm surprised it requires so much vcore to hit 4.6. Maybe run different tests and if the CPU passes without BSOD, gradually lower it a bit and recheck.
Your game crashed, but without the crash details it's tough to pinpoint the cause. It could be an overclock failing under load rather than a hardware issue. Have you tried restoring it to stock briefly to see if the problem persists? It might be a software conflict or another system issue.