Overclock.net › Forums › Intel › Intel CPUs › 5820k degradation at 1.285 volts
New Posts  All Forums:Forum Nav:

5820k degradation at 1.285 volts - Page 6

post #51 of 65
Thread Starter 
you have a better ship than mine,

i lose stability when i put 1.335 volts, meaning 1.325 volts is the max my cpu like, my cpu can't do 4.5 ghz

what did you do to be stable at 4.5 ghz with only 1.310 volts, how much time do you test it ?
post #52 of 65
The funny thing is, I should be even more unstable than you. I have all 8 RAM slots full, and have overclocked my cache frequency to 4.2GHz with 1.25V. I set my LLC to maximum level, and my VCCIN to 1.95V. I have my RAM sticks running at 1.35V, and my VCCSA (system agent) is 1.008V. My CPU freqs are 4.5GHz at 1.310V of Vcore, and I have it 12 hours Realbench stable.

Not to mention that each of my 4GB RAM modules are running at 2666MHz with 13-13-13-30 1T timings. My system stability should be zero theoretically, but my ASUS X99-S is really holding its own. Maybe it's your mobo - you might have to lower cache frequency to 3.0GHz?
Edited by Desolutional - 5/23/15 at 10:53am
post #53 of 65
Thread Starter 
well my cpu realy suck . it one of the worst, if not the worst cpu, the only good thing is my cpu is not so bad at AVX stability

it's like my E8400 EO. it was worst than 99.9% of e8400(EO).

i miss my 2600k , it's was an average cpu, i was able to do 4.7 ghz stable, even 4.8 ghz under 1.4 volts

the only defence my cpu have is 2 r9 290 in crossfire


edit 2 hours stable for now.
Edited by cou51 - 5/23/15 at 11:25am
post #54 of 65
Thread Starter 
my cpu can't be stable at 4.5 ghz, not under 1.35 volts, i think i need more input voltage, like 2.1 volts. it is save to put 2.1V for 24H


my ram is now stable at 13-13-13-30-1t at 2400 mhz with 1.3 volts

post #55 of 65
Quote:
Originally Posted by cou51 View Post

my cpu can't be stable at 4.5 ghz, not under 1.35 volts, i think i need more input voltage, like 2.1 volts. it is save to put 2.1V for 24H


my ram is now stable at 13-13-13-30-1t at 2400 mhz with 1.3 volts


Keep Input Voltage to 2V or lower; higher levels cause degradation. As for not being able to reach 4.5GHz, what cache frequency are you running? Also are you still using ASUS Realbench to stress test?
post #56 of 65
Thread Starter 
yes i use real bench, i can run if for more than 2 hours, but not 8 hours

cache frequency is at 3300 mhz,

i can't get higher than 3300.
post #57 of 65
I don't bother testing longer than 2 hours, any longer than that is unrealistic. In 8 hours, anything can happen, including random errors caused by the idea of entropy. 2 hours should be fine for 24/7 stability.
post #58 of 65
Quote:
Originally Posted by Desolutional View Post

I don't bother testing longer than 2 hours, any longer than that is unrealistic. In 8 hours, anything can happen, including random errors caused by the idea of entropy. 2 hours should be fine for 24/7 stability.


2 hour for stability i says minimum of 12 hour and more if the temp are not crazy


Even after 12 hour my fx did not do error but after like 14 hour it started doing some in p95

And after about 1 week it freezed the system and a week later again upped the voltage redid p95 no error and fully stable



But i may say stress testing those weak intel chip is almost making me feel bad and affraid of blowing it up while i know an amd fx can take a beating
Gaming PC
(18 items)
 
HTPC
(13 items)
 
 
CPUMotherboardGraphicsRAM
Amd Fx 8300 95w Gigabyte 970A-UD3P 8 Phase Sapphire 7950 G.SKILL Trident X DDR3 1600 7-8-8-24 1x8gb 
Hard DriveCoolingOSMonitor
Seagate Barracuda ST1000DM003 1tb Corsair H110i GTX Windows 10 64bit Pro 1440x900 Acer 
KeyboardPowerCaseMouse
Coolermaster Devastator Red Corsair RM650 Coolermaster Haf XB Coolermaster Devastator 
Mouse Pad
E-Blue COBFACE 
  hide details  
Reply
Gaming PC
(18 items)
 
HTPC
(13 items)
 
 
CPUMotherboardGraphicsRAM
Amd Fx 8300 95w Gigabyte 970A-UD3P 8 Phase Sapphire 7950 G.SKILL Trident X DDR3 1600 7-8-8-24 1x8gb 
Hard DriveCoolingOSMonitor
Seagate Barracuda ST1000DM003 1tb Corsair H110i GTX Windows 10 64bit Pro 1440x900 Acer 
KeyboardPowerCaseMouse
Coolermaster Devastator Red Corsair RM650 Coolermaster Haf XB Coolermaster Devastator 
Mouse Pad
E-Blue COBFACE 
  hide details  
Reply
post #59 of 65
Quote:
Originally Posted by bonami2 View Post

2 hour for stability i says minimum of 12 hour and more if the temp are not crazy

Even after 12 hour my fx did not do error but after like 14 hour it started doing some in p95

And after about 1 week it freezed the system and a week later again upped the voltage redid p95 no error and fully stable

But i may say stress testing those weak intel chip is almost making me feel bad and affraid of blowing it up while i know an amd fx can take a beating

Yes, overclocked systems are not meant to be extremely 24/7 stable. Entropy will always affect chips which are run further beyond specifications. The only way to be absolutely safe is to run a chip at stock, or much closer to stock than OCed settings. Stress test time is a personal opinion, some will say 15 minutes, some will say 3 hours, some will say 365 days. I feel personally, as if 2 hours is an even compromise, of course if I was monitoring a patients vital signs, I'd run stock settings. But nothing is mission critical on overclocked systems (why would you overclock in the first place then, just get a professional Xeon with ECC RAM). Your one week thing is purely anomalous. Anything could have cause your system to freeze at that point - including changes in EMF. If you want to be 24/7 stable, then get a Xeon with ECC RAM and a UPS backup supply. And a £10000 PSU, with extreme ripple reduction. Otherwise, a system should be deemed stable after 12 hours as a sensible stable limit. Of course monitor your temperature too, and make sure you've "broken" the chip in before applying 24/7 settings. That means you've run it at stock, for 24 hours using a stress test.
Edited by Desolutional - 5/25/15 at 3:30pm
post #60 of 65
Quote:
Originally Posted by Desolutional View Post

Yes, overclocked systems are not meant to be extremely 24/7 stable. Entropy will always affect chips which are run further beyond specifications. The only way to be absolutely safe is to run a chip at stock, or much closer to stock than OCed settings. Stress test time is a personal opinion, some will say 15 minutes, some will say 3 hours, some will say 365 days. I feel personally, as if 2 hours is an even compromise, of course if I was monitoring a patients vital signs, I'd run stock settings. But nothing is mission critical on overclocked systems (why would you overclock in the first place then, just get a professional Xeon with ECC RAM). Your one week thing is purely anomalous. Anything could have cause your system to freeze at that point - including changes in EMF. If you want to be 24/7 stable, then get a Xeon with ECC RAM and a UPS backup supply. And a £10000 PSU, with extreme ripple reduction. Otherwise, a system should be deemed stable after 12 hours as a sensible stable limit. Of course monitor your temperature too, and make sure you've "broken" the chip in before applying 24/7 settings. That means you've run it at stock, for 24 hours using a stress test.






Go say that to enthusiasm gamer and video editor with 5960x on water or phase change Some do need overclock and stability tongue.gif


Some cant afford a failure and cant afford high end xeon + a gaming desktop with high singlethread so they stress test for hour and hour
Gaming PC
(18 items)
 
HTPC
(13 items)
 
 
CPUMotherboardGraphicsRAM
Amd Fx 8300 95w Gigabyte 970A-UD3P 8 Phase Sapphire 7950 G.SKILL Trident X DDR3 1600 7-8-8-24 1x8gb 
Hard DriveCoolingOSMonitor
Seagate Barracuda ST1000DM003 1tb Corsair H110i GTX Windows 10 64bit Pro 1440x900 Acer 
KeyboardPowerCaseMouse
Coolermaster Devastator Red Corsair RM650 Coolermaster Haf XB Coolermaster Devastator 
Mouse Pad
E-Blue COBFACE 
  hide details  
Reply
Gaming PC
(18 items)
 
HTPC
(13 items)
 
 
CPUMotherboardGraphicsRAM
Amd Fx 8300 95w Gigabyte 970A-UD3P 8 Phase Sapphire 7950 G.SKILL Trident X DDR3 1600 7-8-8-24 1x8gb 
Hard DriveCoolingOSMonitor
Seagate Barracuda ST1000DM003 1tb Corsair H110i GTX Windows 10 64bit Pro 1440x900 Acer 
KeyboardPowerCaseMouse
Coolermaster Devastator Red Corsair RM650 Coolermaster Haf XB Coolermaster Devastator 
Mouse Pad
E-Blue COBFACE 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › 5820k degradation at 1.285 volts