F5F Stay Refreshed Power Users Overclocking The 8700k is now unstable after three weeks.

The 8700k is now unstable after three weeks.

The 8700k is now unstable after three weeks.

Pages (2): 1 2 Next
D
Dr_Venus
Junior Member
11
06-22-2017, 08:43 PM
#1
So I’m experiencing an unusual issue. I changed my 8700k to 5.1ghz at 1.4 vcore, using an LLC at 5. This setup isn’t always running nonstop—just a few hours a day—and the voltage is adaptive with speed scaling, causing the voltage and clock speed to drop significantly when idle. Everything else operates normally. The memory is set to XMP 3200. The system performed well under stress tests like real bench and Prime95 26.6, cinebench. However, yesterday while playing pubg I encountered multiple crashes in a single game. I tried another game and it still crashed, but no error messages appeared. I re-ran the stress test with real bench and found the system unstable. I then reduced the multiplier to 50 and retried, but it didn’t work. It’s puzzling how quickly the system became unstable after just three weeks. Any suggestions on what might be causing this problem? It seems unlikely the temperatures are high since the CPU stays below 80°C during stress tests, and cooling is adequate with water cooling.
D
Dr_Venus
06-22-2017, 08:43 PM #1

So I’m experiencing an unusual issue. I changed my 8700k to 5.1ghz at 1.4 vcore, using an LLC at 5. This setup isn’t always running nonstop—just a few hours a day—and the voltage is adaptive with speed scaling, causing the voltage and clock speed to drop significantly when idle. Everything else operates normally. The memory is set to XMP 3200. The system performed well under stress tests like real bench and Prime95 26.6, cinebench. However, yesterday while playing pubg I encountered multiple crashes in a single game. I tried another game and it still crashed, but no error messages appeared. I re-ran the stress test with real bench and found the system unstable. I then reduced the multiplier to 50 and retried, but it didn’t work. It’s puzzling how quickly the system became unstable after just three weeks. Any suggestions on what might be causing this problem? It seems unlikely the temperatures are high since the CPU stays below 80°C during stress tests, and cooling is adequate with water cooling.

S
SnazzyMeow
Junior Member
45
06-23-2017, 05:27 AM
#2
Just for fun, switch the offset to adaptive plus offset. This helped my i5 8600k reach stable at 5ghz. It's similar to your setup without hyperthreading. I also have my llc at 6 instead of 5.
Of course, this could only work if you didn't mess up your CPU by delidding it. I'm not that confident, but I'm getting 5ghz.
I should note, I also use an MSI build. It's not as powerful as yours, but it's listed below.
S
SnazzyMeow
06-23-2017, 05:27 AM #2

Just for fun, switch the offset to adaptive plus offset. This helped my i5 8600k reach stable at 5ghz. It's similar to your setup without hyperthreading. I also have my llc at 6 instead of 5.
Of course, this could only work if you didn't mess up your CPU by delidding it. I'm not that confident, but I'm getting 5ghz.
I should note, I also use an MSI build. It's not as powerful as yours, but it's listed below.

1
1234qaz12qaz
Posting Freak
773
06-24-2017, 10:57 PM
#3
Running the CPU at elevated voltages may cause harm, and Intel datasheets suggest a maximum of 1.4V. It's likely the current voltage has altered the CPU internally, leading to instability. Have you attempted resetting it to default settings and stress-testing afterward? If successful, consider lowering the overclock and voltage to 1.35V for safer performance.
1
1234qaz12qaz
06-24-2017, 10:57 PM #3

Running the CPU at elevated voltages may cause harm, and Intel datasheets suggest a maximum of 1.4V. It's likely the current voltage has altered the CPU internally, leading to instability. Have you attempted resetting it to default settings and stress-testing afterward? If successful, consider lowering the overclock and voltage to 1.35V for safer performance.

C
CousinBro_Ryan
Junior Member
12
06-25-2017, 03:04 AM
#4
Running the CPU at elevated voltages may cause damage, and Intel data suggests the ideal maximum is around 1.4V. It's possible the voltage you're using has altered the CPU internally, leading to instability. Have you tried resetting it to factory defaults and stress-testing afterward? If that works, consider lowering the overclock to a more conservative setting (like 1.35V) and see your performance. I've managed to stabilize it at 4.8GHz with 1.35V for about 40 minutes using Realbench. I aim to push back up to 5GHz but won't exceed that. I'm concerned about the PSU—only 600W and several years old, which might be affecting stability.
C
CousinBro_Ryan
06-25-2017, 03:04 AM #4

Running the CPU at elevated voltages may cause damage, and Intel data suggests the ideal maximum is around 1.4V. It's possible the voltage you're using has altered the CPU internally, leading to instability. Have you tried resetting it to factory defaults and stress-testing afterward? If that works, consider lowering the overclock to a more conservative setting (like 1.35V) and see your performance. I've managed to stabilize it at 4.8GHz with 1.35V for about 40 minutes using Realbench. I aim to push back up to 5GHz but won't exceed that. I'm concerned about the PSU—only 600W and several years old, which might be affecting stability.

F
Fear_Berzerk
Member
63
07-01-2017, 08:01 PM
#5
It might be the CPU issue if voltage levels aren't stable. A premium HWMonitor version offers live monitoring, allowing you to track performance drops more effectively.
F
Fear_Berzerk
07-01-2017, 08:01 PM #5

It might be the CPU issue if voltage levels aren't stable. A premium HWMonitor version offers live monitoring, allowing you to track performance drops more effectively.

R
Rikuu_Lemon
Junior Member
6
07-21-2017, 04:24 PM
#6
Use hwinfo to check the temperature, generate a log, and note that the free version may trigger overclocking warnings. Be cautious of excessive overclocking to avoid CPU shutdowns for protection.
R
Rikuu_Lemon
07-21-2017, 04:24 PM #6

Use hwinfo to check the temperature, generate a log, and note that the free version may trigger overclocking warnings. Be cautious of excessive overclocking to avoid CPU shutdowns for protection.

M
minime13579
Junior Member
14
07-21-2017, 09:46 PM
#7
scout_03 :
checking hwinfo will help track the temperature, creating a log. It's free version too; you might want to avoid overclocking and let the CPU shut down to protect itself. Yes, thanks. I'm using hwinfo. No voltage spikes detected (LLC level 5, voltage stays near what I set in BIOS). Temperatures aren't a problem since my CPU is delidded and watercooled.
M
minime13579
07-21-2017, 09:46 PM #7

scout_03 :
checking hwinfo will help track the temperature, creating a log. It's free version too; you might want to avoid overclocking and let the CPU shut down to protect itself. Yes, thanks. I'm using hwinfo. No voltage spikes detected (LLC level 5, voltage stays near what I set in BIOS). Temperatures aren't a problem since my CPU is delidded and watercooled.

D
61
07-21-2017, 10:53 PM
#8
verify the ram voltages and ensure they aren't overheating.
D
Dragonwhites06
07-21-2017, 10:53 PM #8

verify the ram voltages and ensure they aren't overheating.

P
pedro_tkf
Senior Member
643
07-23-2017, 01:41 PM
#9
Yes, thanks. I use hwinfo. There were no voltage spikes (using LLC level 5, voltage stayed very close to the setting in bios). Temperatures aren’t a problem since my CPU is delidded and watercooled.
Apologies, it seems there was a typo in my earlier post—I meant to check the PSU, because if the voltages there fluctuate a lot, it could cause the instability you’re seeing. Also, keep in mind that a delid can damage the system—have you tried running this CPU before delidding it, and did you do so recently?
P
pedro_tkf
07-23-2017, 01:41 PM #9

Yes, thanks. I use hwinfo. There were no voltage spikes (using LLC level 5, voltage stayed very close to the setting in bios). Temperatures aren’t a problem since my CPU is delidded and watercooled.
Apologies, it seems there was a typo in my earlier post—I meant to check the PSU, because if the voltages there fluctuate a lot, it could cause the instability you’re seeing. Also, keep in mind that a delid can damage the system—have you tried running this CPU before delidding it, and did you do so recently?

R
Rhuji
Senior Member
437
07-31-2017, 06:23 AM
#10
diellur :
guesswho120 :
yes, thanks. I use hwinfo. There were no voltage spikes (using LLC level 5, voltage stayed very close to what I set in bios). Temperatures weren’t a problem either, since my CPU is delid and watercooled.
Sorry, it seems there was a typo in my earlier post—I meant to check the PSU. If the voltages there were fluctuating a lot, it could cause the instability you’re seeing. Also, keep in mind that a delid can damage the system—have you tried running this CPU before delidding it, and did you do so recently?
I actually delid it recently. I used a rockit88 delid tool and conductonaut liquid metal. Everything seemed okay. I ran the PC for a few days after the delid without any issues.
R
Rhuji
07-31-2017, 06:23 AM #10

diellur :
guesswho120 :
yes, thanks. I use hwinfo. There were no voltage spikes (using LLC level 5, voltage stayed very close to what I set in bios). Temperatures weren’t a problem either, since my CPU is delid and watercooled.
Sorry, it seems there was a typo in my earlier post—I meant to check the PSU. If the voltages there were fluctuating a lot, it could cause the instability you’re seeing. Also, keep in mind that a delid can damage the system—have you tried running this CPU before delidding it, and did you do so recently?
I actually delid it recently. I used a rockit88 delid tool and conductonaut liquid metal. Everything seemed okay. I ran the PC for a few days after the delid without any issues.

Pages (2): 1 2 Next