So i bought a new h100i and the prolimatech pk3 thermal paste cause i do not like the stock ones, i do not trust them. I open the prime 95 v2664 i think, and after 5 minutes it reached 55 celsius on stock speeds. I was pleased cause with the tock cooler it went straight to throttling so i palyed far cry 3. After a couple of hours my temps were 57 at max. Is this nirmal? Are not the temps a little too high with stock volts and speeds? And the thing that surprises me the most is that far cry's 3 temps are in par with the ones in prime 95, which in theory should be a lot higher cause it puts unreal stress to the cpu. Any opinion would be nice.
Solution
You're very welcome.
I understand that you're aware of the version issue with Prime95, but as many other Forum Members and readers lurking in the background are misinformed and have many misconceptions about temperatures, I always appreciate an appropriate opportunity to get the word out.
CT
I understand that you're aware of the version issue with Prime95, but as many other Forum Members and readers lurking in the background are misinformed and have many misconceptions about temperatures, I always appreciate an appropriate opportunity to get the word out.
CT
- Nov 13, 2006
- 6,381
- 290
- 38,240
ifreestylin
Honorable
- Dec 28, 2012
- 991
- 0
- 11,160
I had it at 4.7GHz but i need a bigger PSU. With two GTX 970's and only a 650w, its cutting it too close.
- Nov 13, 2006
- 6,381
- 290
- 38,240
- Nov 13, 2006
- 6,381
- 290
- 38,240
The consensus of the most informed reviewers, writers and overclockers here at Tom's, as well as those at other high profile computer enthusiast websites strongly discourage users from exceeding 1.300 Core voltage on 22 nanometer processors.
If you're using below ambient cooling such as chilled water or LN2, then higher Core voltages may be tolerated.
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.
CPU's become more susceptible to Electromigration with each Die-shrink, so 22 Nanometer architecture is less tolerant of over-volting. Nevertheless, Vcore settings should not exceed the following:
-> Core 2
1st. Generation 65 Nanometer ... 1.50 Vcore
2nd Generation 45 Nanometer ... 1.40 Vcore
-> Core i
1st. Generation 45 Nanometer ... 1.40 Vcore
2nd Generation 32 Nanometer ... 1.35 Vcore
3rd Generation 22 Nanometer ... 1.30 Vcore
4th Generation 22 Nanometer ... 1.30 Vcore
If you're using below ambient cooling such as chilled water or LN2, then higher Core voltages may be tolerated.
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.
CPU's become more susceptible to Electromigration with each Die-shrink, so 22 Nanometer architecture is less tolerant of over-volting. Nevertheless, Vcore settings should not exceed the following:
-> Core 2
1st. Generation 65 Nanometer ... 1.50 Vcore
2nd Generation 45 Nanometer ... 1.40 Vcore
-> Core i
1st. Generation 45 Nanometer ... 1.40 Vcore
2nd Generation 32 Nanometer ... 1.35 Vcore
3rd Generation 22 Nanometer ... 1.30 Vcore
4th Generation 22 Nanometer ... 1.30 Vcore
ifreestylin
Honorable
- Dec 28, 2012
- 991
- 0
- 11,160
- Nov 13, 2006
- 6,381
- 290
- 38,240
It's best to keep this in perspective.
Pushing an exceedingly high Core voltage is simply not worth the risk of degrading your processor for want of another 100MHz. The difference between 4.5 and 4.6 is only 2.2%, which is negligible.
Pushing an exceedingly high Core voltage is simply not worth the risk of degrading your processor for want of another 100MHz. The difference between 4.5 and 4.6 is only 2.2%, which is negligible.
I want to ask you this; 100mhz improves the performance for 2.2%. How much does it affect the lifespan? If the chip was to last 5 years, will it now last 4 and a half? Is there a sure way to know? Because if 0.5 or 1.0 more volts mean that i will get 300 more mhz ( from 4.5 to 4.8) and the incease in performance is ~10% and the lifespan of my chip shortens only by 6 months, i think it worths the risk.
- Nov 13, 2006
- 6,381
- 290
- 38,240
panozy,
The voltage increases you propose will kill your processor, not to mention that your temperatures will be unmanageable.
Once again ... I repeat:
The consensus of the most informed professional reviewers, writers and overclockers here at Tom's, as well as those at other high profile computer enthusiast websites strongly discourage users from exceeding 1.300 Core voltage on 22 nanometer processors.
Respectfully, it's obvious that you do not have a fundamental grasp on overclocking, nor do you understand the relationship between voltage, temperature and watts. I can't condone such a reckless approach. You're on the fast track to destroying your processor within moments, hours, days or weeks.
Clearly you don't understand or appreciate the significance of what I've posted. You need to read, study and comprehend several overclocking guides before you proceed. Here's a good Guide to start with: 3 Step Guide to Overclock Your i7 / i5 Haswell Platform - http://www.overclockers.com/3step-guide-to-overclock-intel-haswell/
I've been building and overclocking computers for over 20 years. The only responsible advice that I can give you is very black and white; stay within recommended limits. This means that you should not exceed 1.300 Core voltage and 80C Core temperature.
This is the best common sense advice you'll find anywhere. Please do not ignore it.
Good luck with your overclocking endeavors.
CT
The voltage increases you propose will kill your processor, not to mention that your temperatures will be unmanageable.
Once again ... I repeat:
The consensus of the most informed professional reviewers, writers and overclockers here at Tom's, as well as those at other high profile computer enthusiast websites strongly discourage users from exceeding 1.300 Core voltage on 22 nanometer processors.
Respectfully, it's obvious that you do not have a fundamental grasp on overclocking, nor do you understand the relationship between voltage, temperature and watts. I can't condone such a reckless approach. You're on the fast track to destroying your processor within moments, hours, days or weeks.
Clearly you don't understand or appreciate the significance of what I've posted. You need to read, study and comprehend several overclocking guides before you proceed. Here's a good Guide to start with: 3 Step Guide to Overclock Your i7 / i5 Haswell Platform - http://www.overclockers.com/3step-guide-to-overclock-intel-haswell/
I've been building and overclocking computers for over 20 years. The only responsible advice that I can give you is very black and white; stay within recommended limits. This means that you should not exceed 1.300 Core voltage and 80C Core temperature.
This is the best common sense advice you'll find anywhere. Please do not ignore it.
Good luck with your overclocking endeavors.
CT
I don't think anyone can give you reliable estimates on something like that. But if you do try, do let us know how long your processor lasts. I don't think it's worth the risk though. Also, from what I've read, you can't cool abve 1.3 Vcore anyway. So there's no point.
CompuTronix :
panozy,
The voltage increases you propose will kill your processor, not to mention that your temperatures will be unmanageable.
Once again ... I repeat:
The consensus of the most informed professional reviewers, writers and overclockers here at Tom's, as well as those at other high profile computer enthusiast websites strongly discourage users from exceeding 1.300 Core voltage on 22 nanometer processors.
Respectfully, it's obvious that you do not have a fundamental grasp on overclocking, nor do you understand the relationship between voltage, temperature and watts. I can't condone such a reckless approach. You're on the fast track to destroying your processor within moments, hours, days or weeks.
Clearly you don't understand or appreciate the significance of what I've posted. You need to read, study and comprehend several overclocking guides before you proceed. Here's a good Guide to start with: 3 Step Guide to Overclock Your i7 / i5 Haswell Platform - http://www.overclockers.com/3step-guide-to-overclock-intel-haswell/
I've been building and overclocking computers for over 20 years. The only responsible advice that I can give you is very black and white; stay within recommended limits. This means that you should not exceed 1.300 Core voltage and 80C Core temperature.
This is the best common sense advice you'll find anywhere. Please do not ignore it.
Good luck with your overclocking endeavors.
CT
The voltage increases you propose will kill your processor, not to mention that your temperatures will be unmanageable.
Once again ... I repeat:
The consensus of the most informed professional reviewers, writers and overclockers here at Tom's, as well as those at other high profile computer enthusiast websites strongly discourage users from exceeding 1.300 Core voltage on 22 nanometer processors.
Respectfully, it's obvious that you do not have a fundamental grasp on overclocking, nor do you understand the relationship between voltage, temperature and watts. I can't condone such a reckless approach. You're on the fast track to destroying your processor within moments, hours, days or weeks.
Clearly you don't understand or appreciate the significance of what I've posted. You need to read, study and comprehend several overclocking guides before you proceed. Here's a good Guide to start with: 3 Step Guide to Overclock Your i7 / i5 Haswell Platform - http://www.overclockers.com/3step-guide-to-overclock-intel-haswell/
I've been building and overclocking computers for over 20 years. The only responsible advice that I can give you is very black and white; stay within recommended limits. This means that you should not exceed 1.300 Core voltage and 80C Core temperature.
This is the best common sense advice you'll find anywhere. Please do not ignore it.
Good luck with your overclocking endeavors.
CT
I do understand, but i was asking for educational purposes. For example, have seen SingularityComputers on youtube? He builds systems for people, but on all of his builds he gives them overclocked cpus with voltages ranging from 1.35 up to 1.4+. I mean, why a profesional would take that risk in his builds? I do not have infinite money to destroying my cpus and then replace and i want to keep mine for at least 3 years. Because i like the conversation we have i want to "extract" as much info i can. Thank you for keep replying.
- Nov 13, 2006
- 6,381
- 290
- 38,240
If you're referring to Singularity Computers Guides, Overclocking, Testing - https://www.youtube.com/playlist?list=PLA6766F5C049874C5 - then you've taken this out of context.
Apples can not be compared to oranges.
The discussions in this thread concern your i7 4790K, which is a 22 nanometer processor.
The overclocked processors shown in the videos in the link above, including Singularity's own personal 3930K rig, are all 32 nanometer processors using extremely high end custom liquid cooling, which is far from what is typical for our readers.
Once again ... from one of my previous posts above:
CPU's become more susceptible to Electromigration with each Die-shrink, so 22 Nanometer architecture is less tolerant of over-volting. Nevertheless, Vcore settings should not exceed the following:
-> Core 2
1st. Generation 65 Nanometer ... 1.50 Vcore
2nd Generation 45 Nanometer ... 1.40 Vcore
-> Core i
1st. Generation 45 Nanometer ... 1.40 Vcore
2nd Generation 32 Nanometer ... 1.35 Vcore <-- i7 3930K
3rd Generation 22 Nanometer ... 1.30 Vcore
4th Generation 22 Nanometer ... 1.30 Vcore <-- i7 4790K
CT
CompuTronix :
If you're referring to Singularity Computers Guides, Overclocking, Testing - https://www.youtube.com/playlist?list=PLA6766F5C049874C5 - then you've taken this out of context.
Apples can not be compared to oranges.
The discussions in this thread concern your i7 4790K, which is a 22 nanometer processor.
The overclocked processors shown in the videos in the link above, including Singularity's own personal 3930K rig, are all 32 nanometer processors using extremely high end custom liquid cooling, which is far from what is typical for our readers.
Once again ... from one of my previous posts above:
CPU's become more susceptible to Electromigration with each Die-shrink, so 22 Nanometer architecture is less tolerant of over-volting. Nevertheless, Vcore settings should not exceed the following:
-> Core 2
1st. Generation 65 Nanometer ... 1.50 Vcore
2nd Generation 45 Nanometer ... 1.40 Vcore
-> Core i
1st. Generation 45 Nanometer ... 1.40 Vcore
2nd Generation 32 Nanometer ... 1.35 Vcore <-- i7 3930K
3rd Generation 22 Nanometer ... 1.30 Vcore
4th Generation 22 Nanometer ... 1.30 Vcore <-- i7 4790K
CT
Yeah, i know that, but if you check his client build 15 part 3, you will see that he overclocked a 4770k with 1.4 volts, that is why i mentioned singularity computers.
- Nov 13, 2006
- 6,381
- 290
- 38,240
Yes, I saw that video as well, however, I can't overemphasize enough that his cooling solutions are far above the practical means and level of technical expertise of almost everyone on these Forums, and is the exception rather than the norm.
All of his work is extremely high-end, and his attention to detail is impeccable, which I greatly appreciate and admire.
Regardless, he overclocked Build 15's i7 4770K to 4.6 at 1.4 on extremely high-end liquid at 68C with 20C ambient, which is 2C below normal or Standard Ambient 22C.
That would make his corrected core temperature 70C.
My de-lidded and lapped 4770K is overclocked to 4.7 at 1.28 on lapped high end air at 78C with Standard Ambient 22C.
I'm under 1.3 volts and 80C, which are the maximum recommended limits.
All of his work is extremely high-end, and his attention to detail is impeccable, which I greatly appreciate and admire.
Regardless, he overclocked Build 15's i7 4770K to 4.6 at 1.4 on extremely high-end liquid at 68C with 20C ambient, which is 2C below normal or Standard Ambient 22C.
That would make his corrected core temperature 70C.
My de-lidded and lapped 4770K is overclocked to 4.7 at 1.28 on lapped high end air at 78C with Standard Ambient 22C.
I'm under 1.3 volts and 80C, which are the maximum recommended limits.
I understand that the difference between a normal custom loop and his loop in huge! But my point was different. Why would he give 1.0 volt more on the cpu if this meant degradation of it=shortened lifespan for a build he was paid to do. He would not risk it, i think. He states that all the time, that he wants to be 100% sure he gives a 24/7 stable system ( after waching all of his videos i came to a conclusion that all the cpus he has been using are not that great, only a couple of them would overclock good even with his awesome wc loops, not that it is his fault, just an observation ). He also said once or twice that his ambient temps are ~30 maybe more, which makes his corrected temps jump to 80+ (not correcting you, it happens that i remember him saying it)
- Nov 13, 2006
- 6,381
- 290
- 38,240
1.0 volts would make it 2.3 ... Incorrect.
0.1 volts would make it 1.4 ... Correct.
Testing close to 22C Ambient is preferred so as to provide normal thermal headroom, but is not required. During the summer season, if adequate A/C is unavailable, then test late at night or early in the morning when Ambient is lowest.
When performing a thermal test below or above 22C, remember to add or subtract the difference so that test results are corrected to Standard Ambient. Here's a summer example:
If measured Ambient is 25C, and reported Core temperature is 80C,
then at standard Ambient 22C, corrected Core temperature would be 77C.
This eliminates variables so results will be consistent and repeatable. Here's a winter example:
If measured Ambient is 19C, and reported Core temperature is 74C,
then at standard Ambient 22C, corrected Core temperature would be 77C.
Again, Standard Ambient temperature is 22C, which is normal room temperature, and is the reference value for Intel’s Thermal Specifications.
I can't account for his willingness to run exceedingly high core voltages, except perhaps he feels it's not a significant problem if you keep the load temperatures low enough.
Unless you have a cooling system of similar capabilities, you will not be able to maintain appropriate load temperatures at that voltage.
However, if you're convinced that it justifies running exceedingly high voltages on your processor, then the responsibility for the health and longevity of your processor is yours alone. Nonetheless, I must still encourage you to be more conservative.
CT
CompuTronix :
1.0 volts would make it 2.3 ... Incorrect.
0.1 volts would make it 1.4 ... Correct.
Testing close to 22C Ambient is preferred so as to provide normal thermal headroom, but is not required. During the summer season, if adequate A/C is unavailable, then test late at night or early in the morning when Ambient is lowest.
When performing a thermal test below or above 22C, remember to add or subtract the difference so that test results are corrected to Standard Ambient. Here's a summer example:
If measured Ambient is 25C, and reported Core temperature is 80C,
then at standard Ambient 22C, corrected Core temperature would be 77C.
This eliminates variables so results will be consistent and repeatable. Here's a winter example:
If measured Ambient is 19C, and reported Core temperature is 74C,
then at standard Ambient 22C, corrected Core temperature would be 77C.
Again, Standard Ambient temperature is 22C, which is normal room temperature, and is the reference value for Intel’s Thermal Specifications.
I can't account for his willingness to run exceedingly high core voltages, except perhaps he feels it's not a significant problem if you keep the load temperatures low enough.
Unless you have a cooling system of similar capabilities, you will not be able to maintain appropriate load temperatures at that voltage.
However, if you're convinced that it justifies running exceedingly high voltages on your processor, then the responsibility for the health and longevity of your processor is yours alone. Nonetheless, I must still encourage you to be more conservative.
CT
I am still not convinced to put so much juice on my cpu, mainly because i do not have enough money to replace it. If i did i would deffinetly try upping the voltages at that point and see the results. My main point for giving singularity as an example, was that he (as a proffesional who is being paid to build a pc and if he makes something wrong he will have to pay the shipment which clearly no one want and few would afford because it is bad for their business,) gave to 4770k 1.4 volts and he thought that it was normal.
4Ryan6
Champion
- Apr 14, 2002
- 20,603
- 7
- 52,915
Hi panozy, First of all IMO anyone that would overclock a computer, any computer, to sell to the public has a screw loose in the brain, overclocking is not finite, many things can affect an overclocks stability to have to be readjusted later.
So when you view this individual as a professional, I view him as psychotic, and I do overclock, but I don't even overclock my own wifes machine, and never one I sold or gave away to a friend, because if something goes wrong with the overclock and it needs some BIOS readjusting, I'm not investing my time to sort it out, even in my own home.
There's a big difference between overclocking for some numbers and bragging rights, vs. overclocking longevity and long term stability and most of what I saw linked in that singularity link were Sandy Bridge 2nd generation Intel CPUs anyway, which could take higher voltages.
No matter what the CPU that's being overclocked, the cooling is what will allow fudging voltages past what is generally safely recommended, which is what CompuTronix is attempting to share with you, staying on the safety side of things.
The safety side is what the overclocking community has found to be safe, and safe being for the longevity of the overclocked CPU, but no one can tell you how long longevity actually is (time wise) with any overclock, simply because the CPUs have not been available long enough to know that.
We always know what happens when too much voltage is applied to a CPU with improper cooling, as it's demise can be instantaneous if the voltage is too high, that's a given!
Voltage increases for longevity of the CPU are suggested to stay on the safe side of the CPUs voltage handling capability, once again the better the cooling the further the voltage can be fudged, and every cooling solution has it's limitations.
I hope this helps your understanding.
4Ryan6 :
Hi panozy, First of all IMO anyone that would overclock a computer, any computer, to sell to the public has a screw loose in the brain, overclocking is not finite, many things can affect an overclocks stability to have to be readjusted later.
So when you view this individual as a professional, I view him as psychotic, and I do overclock, but I don't even overclock my own wifes machine, and never one I sold or gave away to a friend, because if something goes wrong with the overclock and it needs some BIOS readjusting, I'm not investing my time to sort it out, even in my own home.
There's a big difference between overclocking for some numbers and bragging rights, vs. overclocking longevity and long term stability and most of what I saw linked in that singularity link were Sandy Bridge 2nd generation Intel CPUs anyway, which could take higher voltages.
No matter what the CPU that's being overclocked, the cooling is what will allow fudging voltages past what is generally safely recommended, which is what CompuTronix is attempting to share with you, staying on the safety side of things.
The safety side is what the overclocking community has found to be safe, and safe being for the longevity of the overclocked CPU, but no one can tell you how long longevity actually is (time wise) with any overclock, simply because the CPUs have not been available long enough to know that.
We always know what happens when too much voltage is applied to a CPU with improper cooling, as it's demise can be instantaneous if the voltage is too high, that's a given!
Voltage increases for longevity of the CPU are suggested to stay on the safe side of the CPUs voltage handling capability, once again the better the cooling the further the voltage can be fudged, and every cooling solution has it's limitations.
I hope this helps your understanding.
Hehehe a like your first 2 paragraphs, you made me laugh! I like the way you are thinking and i agree with you. This is more of a chat than a thread, i was not really looking for an answer, i was looking for knowledge which all of you gave to me. The purpose of this thread is only educational!
Now, about singularity computers. His work is awesome. He modifies systems, he builds AWESOME wc loops. However, as i said before, he earns his living from that. He does not afford to make a mistake and again for shippment. If you watch his videos closely he always take more precautions than needed. He even uses towel when he bleeds the system! If you check Ronsanut, he does not even care. Singularity is extremely meticulous. He stress tests his systems for 72 hours and if the client requests OC he puts another 72h of testing! OC is something everyone can do, so instead of doing it themselves, the clients request it and he does, there is no difference. And once again, because he lives from building pc's, he doesnot afford to lose time and money because he was careless. He even provides a 2year warranty!
Anyway, back to the volt, oc anx temp thing. I know that every cooling solution has limits. I know that cpu's have limits. But i am trying to find what would happen in every theoretical situation. 0.1 volts would make such a difference in the lifespan of a cpu? For instance, you have 4.7ghz@1.285 volts and for 4.8 you need exactly 1.31 vllta, with temps <75. How this would affect the cpu in the longrun?
4Ryan6
Champion
- Apr 14, 2002
- 20,603
- 7
- 52,915
If that's the case you probably should have posted the thread as a discussion and not a question.
Regarding the singularity computers his work is as common as can be there's nothing special about what he does, to the newbie it may come across as awesome, but to those water coolers and case modders that know what they're doing, he is old school at best.
Shipment costs are the least of his concerns with a failed overclocked setups, lawsuit maybe!, but shipment, Nah!
He doesn't seem to be listed at the Better Business Bureau? Hmm? Wonder Why?
Regarding your last paragraph of voltage increases and CPU longevity I've already answered that.
TRENDING THREADS
-
-
News Introducing the Tom’s Hardware Premium Beta: Exclusive content for members
- Started by Admin
- Replies: 40
-
-
-
-
-
Discussion What's your favourite video game you've been playing?- Started by amdfangirl
- Replies: 4K
Latest posts
-
Question I upgraded my power supply and now i cant get any display to come through.
- Latest: taylo_galbraith
-
-
-
-
-
-
Facebook
Twitter
Instagram