Same main issue persists with Prime95 testing on this 6600k OC MSI M5 gaming PC with NH-C14 cooling.
Same main issue persists with Prime95 testing on this 6600k OC MSI M5 gaming PC with NH-C14 cooling.
I was attempting to optimize the OC settings for my i5 6600k yesterday, aiming for a good balance between performance and temperature.
No matter the combinations I tried, prime95 consistently showed the same core failing much sooner than others.
Last night I ran prime95 and then slept; upon waking, I noticed three cores still running, one that stopped after five minutes of testing.
Temperatures stayed around 70-75°C.
The settings I used were: 4.5GHz, 100BLKC speed, 2800dram (from 2600), 1.4v for CPU, 1.32 for dram.
Could anyone assist in figuring out the best course of action?
Does this indicate a flaw in the core that consistently fails?
Should I consider an OC for each core individually—would that be feasible?
EDIT – if necessary, I can share test results and screenshots when I get back from work.
1.4v is too high! The maximum you should use is 1.35v. Using higher speeds will damage your chip. Some i5 models can't run at 4.5; it's a matter of chance. Try 4.2ghz @ 1.3v first, then adjust to 4.3ghz if it meets p95, and finally 4.4. Intel temperature guide:
1.4v is too high! The maximum you should use is 1.35v. Using higher speeds will damage your chip. Some i5 models can't run at 4.5; it's a matter of chance. Try 4.2ghz @ 1.3v first, then adjust to 4.3ghz if it meets p95, and finally 4.4. Intel temperature guide:
N3rdR4ge: 1.4v is excessive! The maximum safe value is 1.35v. Using 1.4v will likely damage your chip. Some i5 models struggle with 4.5, which depends on the specific situation. Adjust to 4.2ghz @ 1.3v and observe the results. Increase to 4.3ghz if it meets p95, then 4.4. Intel temperature guide: Thanks for your response. Is 1.4v really too high? Is it the voltage or the temperature that causes issues? If I could run at 1.5 volts and still keep temperatures under 80, wouldn’t that be safe? Also, regarding your comment, I’ve tried 4.4ghz with 1.38v and the test is currently running at home. I’ll update you once I’m back from work. As mentioned before, at 4.5ghz and 1.4volt, all three cores passed the test smoothly throughout the night without issues. The fourth core only failed after nearly five minutes of testing. This is what concerns me the most. The three processors appear capable of handling even higher speeds than 4.5ghz easily.
if a core fails you are pushing it to hard. doesnt matter if 3/4 cores can handle it. all or nothing. it is common to have one core fail much quicker/easier than others. nothing to be concerned about. 1.35 is the recommended maximum, otherwise you will cause your chip to have a reduced lifespan.
please review the temp guide i linked to you...
"Excessive Vcore and temperatures will result in accelerated "Electromigration" -
https://www.google.com/?gws_rd=ssl#q=Electromigration
- which prematurely erodes the traces and junctions within the processor's layers and nano-circuits. This will eventually result in blue-screen crashes, which will become increasingly frequent over time."
N3rdR4ge:
if a core fails you are pushing it to hard. doesn't matter if 3/4 cores can handle it. all or nothing. it is common to have one core fail much quicker/easier than others. nothing to be concerned about. 1.35 is the recommended maximum, otherwise you will cause your chip to have a reduced lifespan.
please review the temp guide i linked to you...
"Excessive Vcore and temperatures will result in accelerated "Electromigration" -
https://www.google.com/?gws_rd=ssl#q=Electromigration
- which prematurely erodes the traces and junctions within the processor's layers and nano-circuits. This will eventually result in blue-screen crashes, which will become increasingly frequent over time."
I will read it tonight. thank you.
however, can you please tell me why i should worry about the volt and not the temperature?
I really do not understand why having higher voltage with OK temperatures is bad.
So u are saying the max volt i should go for my 6600k is 1.35? and using that voltage i should aim for 4.2 maybe 4.3. Right?
By the way, is there an option to OC each core individually?
meaning that i push more the three that work fine, and keep the "defective" one with a lower multiplier?
It looks like the OC at 4.4GHz passed the prime95 test. It operated for about 9 hours without any issues.
Screenshot: http://picpaste.com/cpu_benchmark-95utdJFW.jpg
The maximum temperature reached was 76°C, which is the highest I've seen so far.
Idle temperature stayed around 30°C.
EDIT: This high temperature came after many hours of full testing. At first it was about 60°C.
The settings I used were:
CPU ratio 44 (originally 35)
CPU base Clock: 100 MHZ (original)
XMP enabled: 2800MHZ (originally 2666)
CPU Core Voltage: 1.38V
DRAM Voltage: 1.35V
I’m wondering if I should stick with these settings or lower the OC because the temperature reached 76°C after all that testing.
Please, experts, let me know!! 😀
I clarified what high voltages do. I even cited the site. It's called "Electromigration," and you should look it up. I included the link from my previous post with the quote. Temperature doesn't matter if you're applying too many volts to your chip. Try 1.35 with 4.4, then drop it to 4.3 if it fails.
Make sure p95 you run a smallFFT.
I suggest you stop overclocking your DRAM and DRAM voltage first. This is likely the reason for your failures. Focus on CPU overclocking before worrying about DRAM. 2666 is sufficient speed anyway. Overclocking RAM won't provide much benefit at all.
I previously achieved 4.4ghz with my i5-4690k at 1.21v. 1.38v is significantly higher and may not be compatible if the RAM doesn't work. It's a matter of luck. Check the temperature guide.
I previously achieved a performance of 4.4ghz with my i5-4690k running at 1.21v. 1.38v would be significantly excessive. It seems the processor might not be suitable if the RAM isn’t functioning correctly. This outcome depends heavily on luck, and it’s important to follow the temperature guidelines.
I regret bringing this up again, but I share a nearly identical issue with the original poster, using the same motherboard and processor. It could be useful for you.
My attempts have been to push my CPU overclocking to 4.3–4.5ghz and voltages between 1.25–1.36v. Each time, I successfully completed Small FFT tests lasting over 12 hours with temperatures under 70°C, but the Blend test consistently fails the same core while others remain stable for extended periods afterward. This inconsistency is particularly irritating because the main variation lies in RAM, yet I haven’t adjusted it.
Could this be a risk if I disregard the Blend results and continue using the system normally? Most of my usage is gaming, so I’m not concerned about stability.
Below are my specifications:
PCPartPicker part list
Price details by seller