Help me OC my PC
Help me OC my PC
Looking for guidance on OC to boost performance in games? This is the most frequent reason people ask. You want tips without relying on Skype or similar platforms. Your setup includes a variety of components: an i7 6700k, Gigabyte 1060 6 GB Xtreme, ASUS Z170 Pro, 8 GB of 2133 MHz RAM (two 8 GB modules), EVGA SuperNova 750W G2, WD Blue 1TB HDD, Fractal Design R5, MSI afterburner and Rivatuner, and a UniGine Valley for benchmarking. You're gaming at 1080p on two monitors—one dedicated GPU, the other integrated for web use.
With a powerful CPU, you likely won't notice any improvement in gaming performance from overclocking. However, if you're determined, you can achieve a basic overclock by adjusting settings. Set your CPU core voltage to 1.2V and the multiplier to 40. Ensure your computer runs games smoothly without crashes and stays below 75°C. If it functions properly, increase the multiplier to 41. Continue this process until it stops working or overheats. Adjust the multiplier slightly lower when it fails. You've successfully overclocked your CPU.
With adequate cooling, a 6700k should handle 4.6 without overheating... Lodders offered useful tips on managing it... Just rinse and repeat, increasing the multiplier until it stops climbing further... You might be able to boost the voltage, but I tend to avoid that since it often leads to lower returns... For instance, my 6700k is at 4.6 with 1.23v, stable at 5.0 with 1.43v, but temperatures sit in the 70s and low 80s. At 4.6 I’m seeing a cool 50°C under load and mid-60s at full power. The gap between 4.6 and 5.0 doesn’t seem worth the extra voltage or heat...
With a CPU that's so good, you likely won't notice any improvement in gaming performance from overclocking. However, if you're determined, you can try a basic overclock by adjusting your CPU core voltage to 1.2V and setting the multiplier to 40. Make sure your computer runs games smoothly without crashes and stays below 75°C. If it works, increase the multiplier until it fails or overheats, then reduce it slightly. Congratulations, you've successfully overclocked your CPU.
Erothes :
Makentox :
What kind of CPU cooler do you use?
I have a TX3 Evo cooler, but I think I managed to get a fairly good one.
I won’t ask why I bought such a subpar cooler—it’s not ideal for overclocking—but I’m sure you can still push your CPU up to around 4.5 or 4.6 with acceptable temps. It really depends on the quality of your chip and the airflow in your case.
Download prime95 for stress testing and use HWmonitor to keep an eye on temperatures.
Set manual voltage to 1.300V, LLC to level 6 (you’ll find this in the VRM settings), and adjust the multiplier to 46. Run a stress test. If it stays stable, go back to BIOS and lower the voltage by 0.010V, then repeat until the system starts acting unstable. Once that happens, increase the voltage by another 0.010V. This voltage passed the stress test but doesn’t guarantee stability in games. From now on, just play games; if you notice a further rise of 0.010V, it should be fine.
For the GPU, it’s straightforward—keep increasing the clock speed until the GPU starts crashing. Then perform a memory stress test using FurMark and watch for any artifacts.
From my experience, these are the most demanding games! CPU rises in Tomb Raider is impressive, while Witcher3, Metro, and Crysis pale in comparison.
For GPU, Rise of Tomb Raider really pushed memory limits, and the free Steam game "Unturned" significantly affected GPU clock speeds.
Makentox shared tips on checking your CPU cooler and monitoring temperatures. He recommends using Prime95 for stress testing and HWmonitor for tracking temps. He suggests adjusting voltage manually, setting LLC to level 6, and using a multiplier of 46. He advises repeating the test until stability is achieved, then gradually increasing voltage if needed. For GPU, he advises pushing clock speeds until crashes occur and running memory stress tests with FurMark to ensure no artifacts appear. He also mentions popular games for testing, such as Tomb Raider, Witcher3, Metro, and Crysis, noting their high demands. He asks about the duration of stress tests and clarifies questions on LLC settings and stability indicators.
Erothes :
Makentox :
Erothes :
Makentox :
What about your CPU cooler?
My CPU cooler is a TX3 EVO, but I think I managed to get a fairly good one. I won’t ask why I bought such a mediocre cooler—it’s not suitable for overclocking—but I’m sure you can still push your CPU up to around 4.5 or 4.6 with decent temps. It really depends on the quality of your chip and the airflow in your case.
Download prime95 for stress testing and use HWmonitor to keep an eye on temperatures.
Set manual voltage to 1.300V and LLC to level 6 (you’ll find this in the VRM settings), and adjust the multiplier to 46. Run a stress test on your system. If it stays stable, go back to BIOS and lower the voltage by 0.010V, repeat until the system starts acting unstable. Once that happens, increase the voltage by another 0.010V. This voltage passed the prime95 test but doesn’t guarantee stability in games. From then on, just play games; if you notice a further rise of 0.010V, it should be fine.
For the GPU, it’s quite straightforward—keep increasing your clock speed until the GPU starts crashing, then repeat with memory stress tests using FurMark, making sure no artifacts appear.
From my experience, these are the best games to test! CPU rise in Tomb Raider is intense; no game is more demanding than this. Witcher3, Metro, and Crysis pale in comparison.
For GPU rise in Tomb Raider, memory had a major impact, and Steam games like "Unturned" significantly affected GPU clock speeds.
How long should I run the stress test?
10 to 15 minutes
What is LLC and why should I set it to level 6?
Before LLC was introduced, overclocking involved dealing with a nasty problem called Vdroop. Vdroop leads to a reduction in CPU voltage as the load increases. The system can’t keep the set VCore stable, which is essential for maintaining the OC setting. Under load, CPU voltage would drop so much that frequent BSODs or crashes would happen. It was especially problematic when you believed you had found the perfect 24/7 OC—Vdroop would ruin the experience by making your system unstable.
How do I know if it’s unstable?
BSOD
You should definitely monitor your temperature during stress testing and avoid letting it exceed 80°C.
Makentox :
Erothes :
Makentox :
Erothes :
Makentox :
What about your CPU cooler?
My CPU cooler is a TX3 EVO, but I think I managed to get a fairly good one. I won’t ask why I bought such a mediocre cooler—it’s not suitable for overclocking—but I’m sure you can still push your CPU up to around 4.5 or 4.6 with decent temps. It really depends on the quality of your chip and the airflow in your case.
Download prime95 for stress testing and use HWmonitor to keep an eye on temperatures.
Set manual voltage to 1.300V, LLC to level 6 (you’ll find this in the VRM settings), and adjust the multiplier to 46. Run a stress test. If it stays stable, go back to BIOS and lower the voltage by 0.010V, repeat until the system starts acting up. Once that happens, increase the voltage by 0.010V again. This voltage passed prime95, but it doesn’t guarantee stability in games. From now on, just play games; if you notice a further rise of 0.010V, it should be fine.
For the GPU, it’s pretty straightforward—boost your clock speed until the GPU starts failing, then repeat with memory stress tests using FurMark, making sure no artifacts appear.
From my experience, these are the most demanding games! CPU rise in Tomb Raider is intense; no game compares. Witcher3, Metro, and Crysis pale in comparison.
For GPU rise in Tomb Raider, memory was a big factor, and Steam games like "Unturned" significantly affected GPU clock speed.
How long should I run the stress test?
10 to 15 minutes
What is LLC and why should I set it to level 6?
Before LLC was introduced, overclocking involved dealing with Vdroop—a problem that causes CPU voltage to drop as load increases. The system can’t keep the desired voltage stable, which is crucial for maintaining a consistent overclock setting. Under heavy load, the CPU voltage would fall so low that frequent BSODs or crashes would occur. This was especially problematic when you believed you had found the perfect constant OC; Vdroop would disrupt everything and ruin the experience.
How do I tell if it’s unstable?
BSOD
You should keep an eye on your temperature during stress tests, keeping it below 80°C.
I’m wondering if I should run the prime95 blend?
And why do both CPUID HWmonitor and CPU-Z show my core voltage around 0.8? When I set it in BIOS to 1.300 and in AI Suite to 3, it still reads about 1.3—does that mean it’s only under load?
Also, does an 80+ temperature mean instability?
And should I keep the minimum multiplier at 8?