my first overclock how did i do

[brandonb21]

 

my cooler is an h80i in push pull 

[Interfectorem]

74 is a bit on the high side I would say, but I've never owned an Intel so I wouldn't know. Looks great. I bet 4.6 GHz is beast.

[L.E.D]

 

my cooler is an h80i in push pull 

 

Temps are a little high for my liking, other then that I'd say it's golden.

[brandonb21]

74 is a bit on the high side I would say, but I've never owned an Intel so I wouldn't know. Looks great. I bet 4.6 GHz is beast.

 

you gotta remember you will never see these temps under a real world senairo anyways

[Krzyimprt]

1.38v is extremly high for 4.6

[brandonb21]

1.38v is extremly high for 4.6

il try leveling it down 

[Interfectorem]

you gotta remember you will never see these temps under a real world senairo anyways

True. But it's always a possibility. Yea, I'm such a wimp on temps that I undervolted my FX-8320 on stock speeds. It stays pretty darn cool with my mediocre 120M thin rad in push pull. Maxing out at  51C during Prime95. However 60C is threshold for these CPUs

 

I'm picking up some extra shifts to pick up a nice NZXT Kraken x61 and a case that actually supports rads up to, because mine doesn't.

[NeatSquidYT]

Mess around with the voltages a bit more...

[Krzyimprt]

il try leveling it down 

Try 4.6ghz @ 1.25v if it's not stable increase in small increments. You have to find what your chip can do, it's the luck of the draw. You might need a huge voltage increase for a 200mhz bump which is not worth it.

[loccilucas]

74 is a bit on the high side I would say, but I've never owned an Intel so I wouldn't know. Looks great. I bet 4.6 GHz is beast.

It isn't too bad, mine maxes out at 80c after about half an hour of p95, Considering it will never get that high in real use its nothing to worry about.

[Faceman]

@brandonb21

 

Being a novice overclocker, I spent a lot of time researching how to overclock my i5-4670k on Asus Z87-A motherboard.  After a lot of trial and error doing it manually, I decided to give the auto-overclocking a try.  Lo-and-Behold, the Auto Overclock set my system to 4.7Ghz @ 1.275v, which is an incredible overclock and a lot more robust than anything I had hoped for or tried to do manually.  After gaining more knowledge and confidence, I revisited my auto-overclock and learned that while it was a good result, some of the secondary settings were not very safe.  Namely Load-Line Calibration.  The auto-OC had set my LLC to the maximum setting, and this was causing massive voltage spikes, beyond the preset 1.275v, which is not good for the processor.  While everything was stable, you want the lowest possible voltage for a given multiplier to keep it from crashing and to eliminate excess heat.

 

     Many people have had very bad results with their auto-OC, with it just not working what-so-ever, and when they go to do it manually, they get it working on their own.  My successful auto-overclock seems to be a rare, but possible result.  There are others who have been successful using Asus' Auto-OC, but this is somewhat rare, and manual is always better.

 

     My advice--Give the auto OC a whirl to get a baseline of how your chip performs. It doesn't take long, and if it doesn't work out well, you can always set it to default and try it manually.  If it does give you a good result, remember the general settings and go back into the BIOS and plug those settings in manually.

 

Before you manually overclock, do some research.  Here are some excellent guides, be aware that these are Haswell and Devil's Canyon Overclocking guides, so the results will likely be different from the 5820k, but at least it gives you an idea.

Asus BIOS Overclocking Terminology

Overclocker's.Net Overclocking Guide

Linus Tech Tips Overclocking Guide

Load Line Calibration and Overclocking

 

     You have to go into overclocking with an open mind.  The goal of overclocking should be to get the best possible results for your chip.  Every chip is unique and your results are going to be different from someone else with the exact same system.  It is all predicated upon the chip or Silicon Lottery as some like to call it.  Don't try to replicate other people's results, focus on getting the best possible overclock for your chip that you are comfortable with.  Make sure that you have proper cooling, especially with such a beastly processor, and the most important rule of overclocking: SET VOLTAGE TO MANUAL BEFORE STRESS TESTING.  If you leave your voltage on adaptive mode before running a stress test, you run the risk of overvolting and destroying/degrading your processor and in the worst case scenario, death of the processor.  I cannot stress this enough, set your voltage to manual before stress testing.  Once you have finished stress testing, then you can revert back to adaptive.

 

     Edit*  You are going to need some programs to monitor your results and to stress test.  The best monitoring program is HWiNFO. RealTemp is also a popular option.  CPU-Z is another very useful program, but the most accurate readings will come from HWiNFO.  For stress testing, you have a lot of options to choose from, some more strenuous than others.

 

Prime95 v27.9 is a very popular option, but you must be careful with this program.  It is arguably the most strenuous stress test and it has the potential to damage your processor if used improperly.  You must always set your voltage to manual before using Prime95.  Failure to do so can and will result in your processor being damaged and potential death of the processor.  Because Prime is so stressful, instability is often found within the first 5min, making it a good "quick and dirty" stress test.  I personally never use P95 for longer than 20min.  Prime95 is going to produce an unrealistic amount of heat on your processor, you must have adequate cooling and the balls to put up with this amount of heat.  Real-world use will never come close to the heat that Prime95 causes.

Here are the Custom Settings you want to use: Threads = (*However many Threads your Processor is capable of*), Min FFT Size(in K) = 8, Max FFT Size(in K) = 4096, Memory to Use(in MB) = 1600, Time to Run Each FFT Size(in Minutes) = 3.

You want to make Time to Run Each Test 3min so that a variety of tests ranging in size will be run in a shorter time span, making the actual testing not be so long.  Lots of people recommend stress testing for hours upon hours, its just not necessary.  If your system isn't stable, it is going to let you know very quickly.

CinebenchR15 is another good and free program.  This is a more "real-world" type test where your CPU renders a simple image in a short amount of time resulting in a score.  The processor is placed under full 100% load during this test.  It is a good idea to run this test 3-5 times to check for stability.  Again, it is a good idea to set your voltage to manual before doing this.  This test is not as strenuous as Prime95, temperatures won't be nearly as high.

AIDA64 Is another popular stress test program.  I personally have never used it, but many people swear by it.  Remember to set your voltage to manual.  This is becoming a recurring theme.  Always set your voltage to manual before stress testing.

Intel Extreme Tuning Utility(XTU) This is a nice program that is more than just a stress test and bench mark, it gives you the option to fiddle with settings while in Windows. It is not recommended to do so, but it can be done.  Be sure to not have more than 1 overclocking program running at once because the values will conflict and chaos will ensue. The stress test in XTU uses a derivative of Prime95, but is not as stressful as the real deal. As always, voltage set to manual before stress testing.

Intel Burn Test I don't have experience with this one personally, but other people recommend it.  Voltage on manual.

OCCT  Never used it, but another option for stress testing.  Manual voltage required.

Linpack Never used, but another option for testing.  Manual Voltage Required.

 

Lastly, play games and do normal tasks.  I have found BF4 and Crysis3 to be very good "stress tests" because it is what I use my machine for-gaming.  It gives me a realistic scenario to test out my overclock.  For games and everything other than stress testing, you should use Adaptive Voltage.

 

One final thing to note:  What may be stable in one program, may not be stable in another.  I remember when I was dialing in my medium overclock I had found that I was stable in Prime95-4.5Ghz @ 1.185v w/ Manual Voltage, but I was not stable in BF4 when at 4.5Ghz @ 1.185v Adaptive.  I had to push my voltage up to 1.200v adaptive @ 4.5Ghz to be stable in BF4.

[brandonb21]

 

 

roound 2 how is it now

[brandonb21]

@brandonb21

 

Being a novice overclocker, I spent a lot of time researching how to overclock my i5-4670k on Asus Z87-A motherboard.  After a lot of trial and error doing it manually, I decided to give the auto-overclocking a try.  Lo-and-Behold, the Auto Overclock set my system to 4.7Ghz @ 1.275v, which is an incredible overclock and a lot more robust than anything I had hoped for or tried to do manually.  After gaining more knowledge and confidence, I revisited my auto-overclock and learned that while it was a good result, some of the secondary settings were not very safe.  Namely Load-Line Calibration.  The auto-OC had set my LLC to the maximum setting, and this was causing massive voltage spikes, beyond the preset 1.275v, which is not good for the processor.  While everything was stable, you want the lowest possible voltage for a given multiplier to keep it from crashing and to eliminate excess heat.

 

     Many people have had very bad results with their auto-OC, with it just not working what-so-ever, and when they go to do it manually, they get it working on their own.  My successful auto-overclock seems to be a rare, but possible result.  There are others who have been successful using Asus' Auto-OC, but this is somewhat rare, and manual is always better.

 

     My advice--Give the auto OC a whirl to get a baseline of how your chip performs. It doesn't take long, and if it doesn't work out well, you can always set it to default and try it manually.  If it does give you a good result, remember the general settings and go back into the BIOS and plug those settings in manually.

 

Before you manually overclock, do some research.  Here are some excellent guides, be aware that these are Haswell and Devil's Canyon Overclocking guides, so the results will likely be different from the 5820k, but at least it gives you an idea.

Asus BIOS Overclocking Terminology

Overclocker's.Net Overclocking Guide

Linus Tech Tips Overclocking Guide

Load Line Calibration and Overclocking

 

     You have to go into overclocking with an open mind.  The goal of overclocking should be to get the best possible results for your chip.  Every chip is unique and your results are going to be different from someone else with the exact same system.  It is all predicated upon the chip or Silicon Lottery as some like to call it.  Don't try to replicate other people's results, focus on getting the best possible overclock for your chip that you are comfortable with.  Make sure that you have proper cooling, especially with such a beastly processor, and the most important rule of overclocking: SET VOLTAGE TO MANUAL BEFORE STRESS TESTING.  If you leave your voltage on adaptive mode before running a stress test, you run the risk of overvolting and destroying/degrading your processor and in the worst case scenario, death of the processor.  I cannot stress this enough, set your voltage to manual before stress testing.  Once you have finished stress testing, then you can revert back to adaptive.

 

     Edit*  You are going to need some programs to monitor your results and to stress test.  The best monitoring program is HWiNFO. RealTemp is also a popular option.  CPU-Z is another very useful program, but the most accurate readings will come from HWiNFO.  For stress testing, you have a lot of options to choose from, some more strenuous than others.

 

Prime95 v27.9 is a very popular option, but you must be careful with this program.  It is arguably the most strenuous stress test and it has the potential to damage your processor if used improperly.  You must always set your voltage to manual before using Prime95.  Failure to do so can and will result in your processor being damaged and potential death of the processor.  Because Prime is so stressful, instability is often found within the first 5min, making it a good "quick and dirty" stress test.  I personally never use P95 for longer than 20min.  Prime95 is going to produce an unrealistic amount of heat on your processor, you must have adequate cooling and the balls to put up with this amount of heat.  Real-world use will never come close to the heat that Prime95 causes.

Here are the Custom Settings you want to use: Threads = (*However many Threads your Processor is capable of*), Min FFT Size(in K) = 8, Max FFT Size(in K) = 4096, Memory to Use(in MB) = 1600, Time to Run Each FFT Size(in Minutes) = 3.

You want to make Time to Run Each Test 3min so that a variety of tests ranging in size will be run in a shorter time span, making the actual testing not be so long.  Lots of people recommend stress testing for hours upon hours, its just not necessary.  If your system isn't stable, it is going to let you know very quickly.

CinebenchR15 is another good and free program.  This is a more "real-world" type test where your CPU renders a simple image in a short amount of time resulting in a score.  The processor is placed under full 100% load during this test.  It is a good idea to run this test 3-5 times to check for stability.  Again, it is a good idea to set your voltage to manual before doing this.  This test is not as strenuous as Prime95, temperatures won't be nearly as high.

AIDA64 Is another popular stress test program.  I personally have never used it, but many people swear by it.  Remember to set your voltage to manual.  This is becoming a recurring theme.  Always set your voltage to manual before stress testing.

Intel Extreme Tuning Utility(XTU) This is a nice program that is more than just a stress test and bench mark, it gives you the option to fiddle with settings while in Windows. It is not recommended to do so, but it can be done.  Be sure to not have more than 1 overclocking program running at once because the values will conflict and chaos will ensue. The stress test in XTU uses a derivative of Prime95, but is not as stressful as the real deal. As always, voltage set to manual before stress testing.

Intel Burn Test I don't have experience with this one personally, but other people recommend it.  Voltage on manual.

OCCT  Never used it, but another option for stress testing.  Manual voltage required.

Linpack Never used, but another option for testing.  Manual Voltage Required.

 

Lastly, play games and do normal tasks.  I have found BF4 and Crysis3 to be very good "stress tests" because it is what I use my machine for-gaming.  It gives me a realistic scenario to test out my overclock.  For games and everything other than stress testing, you should use Adaptive Voltage.

 

One final thing to note:  What may be stable in one program, may not be stable in another.  I remember when I was dialing in my medium overclock I had found that I was stable in Prime95-4.5Ghz @ 1.185v w/ Manual Voltage, but I was not stable in BF4 when at 4.5Ghz @ 1.185v Adaptive.  I had to push my voltage up to 1.200v adaptive @ 4.5Ghz to be stable in BF4.

 see picture above that was the closest i could get it anything below 1.3 volts would bluescreen so i kept dialing it down till i got to 1.301 seems stables 30mins in a burn in on manual voltage

[Faceman]

 see picture above that was the closest i could get it anything below 1.3 volts would bluescreen so i kept dialing it down till i got to 1.301 seems stables 30mins in a burn in on manual voltage

4.7 @ 1.3 is solid for a 4790k.

[brandonb21]

4.7 @ 1.3 is solid for a 4790k.

temps are up their tho 75 78 sometimes 80 but no higher

[Faceman]

temps are up their tho 75 78 sometimes 80 but no higher

Keep it below 85C, and this is a stress test, so real-world won't be as bad.  Make sure that your h80i is mounted properly.

[brandonb21]

Keep it below 85C, and this is a stress test, so real-world won't be as bad.  Make sure that your h80i is mounted properly.

it is i had it off twice already had issues with stock temps due to a loose backplate. i did the washer trick but ran out of thermal paste none is sold anywheres near me so i just had to reuse even so the temps are better, but could be better with artic sliver

[brandonb21]

update

 

it is i had it off twice already had issues with stock temps due to a loose backplate. i did the washer trick but ran out of thermal paste none is sold anywheres near me so i just had to reuse even so the temps are better, but could be better with artic sliver

 

[Faceman]

it is i had it off twice already had issues with stock temps due to a loose backplate. i did the washer trick but ran out of thermal paste none is sold anywheres near me so i just had to reuse even so the temps are better, but could be better with artic sliver

Arctic Silver is not a good brand of thermal paste. It is old, it has a long cure time, and it is conductive.  Much better thermal paste can be purchase for the same price or less.

 

Check out this Thermal Paste Roundup

 

I personally use Gelid GC-Extreme and it works like a charm.  Another very good quality thermal paste that isn't mentioned on there is IC Diamond, if it was on that list, it would be near the top.

[brandonb21]

Arctic Silver is not a good brand of thermal paste. It is old, it has a long cure time, and it is conductive.  Much better thermal paste can be purchase for the same price or less.

 

Check out this Thermal Paste Roundup

 

I personally use Gelid GC-Extreme and it works like a charm.  Another very good quality thermal paste that isn't mentioned on there is IC Diamond, if it was on that list, it would be near the top.

 

how long should i let the stress test go its been going for 45mins how long till i should go into bios and figure out adaptive voltage? 

[Faceman]

how long should i let the stress test go its been going for 45mins how long till i should go into bios and figure out adaptive voltage? 

I never used AIDA before so I can't say for certain.  Your temperatures are high for AIDA64 though, by about 15-20C.

[brandonb21]

I never used AIDA before so I can't say for certain.  Your temperatures are high for AIDA64 though, by about 15-20C.

someone told me for an h80i thoes temps are fine new picture incoming soon. 

[brandonb21]

[Faceman]

someone told me for an h80i thoes temps are fine new picture incoming soon. 

Then the H80i is a complete waste of money because it performs about as well as a CM 212 EVO.  I think yours is not mounted properly, or the thermal paste situation is to blame.  Regardless, I think you're stable in AIDA64.  Go ahead and swap to adaptive voltage and play some games.  BF4 or Crysis3 are good ones because they stress the CPU a considerable amount.

[brandonb21]

Then the H80i is a complete waste of money because it performs about as well as a CM 212 EVO.  I think yours is not mounted properly, or the thermal paste situation is to blame.  Regardless, I think you're stable in AIDA64.  Go ahead and swap to adaptive voltage and play some games.  BF4 or Crysis3 are good ones because they stress the CPU a considerable amount.

yeah il replace the thermal pase or i might replace to the cooler all togother with an h105 in the coming months  i'm 100% sure its the thermal paste the pump is mounted on their very good i even did the washer trick