Overclocking the i7 4790k

[Borge80]

After reading reviews of people managing stable clocks of 4.8ghz, I managed to get my cpu to 4.8ghz with a manual 1.375vcore (system bumps it to 1.392) and idles on windows fine. I stress tested it for 1-2 minutes at 100% load and was hitting 100c max and it throttled, still did not crash. Anything lower than a 1.375 manual vcore crashes instantly. Tried 4.9ghz and 5ghz as well and was working with the voltages for some time, but cannot find the magic here. Insta-crashes at all voltages. 5ghz leaves it at 1.45 vcore which is uncomfortably high.

Idle temps are 28c with sunbeam core contact freezer as a heat sink.
Then I loaded up Far Cry 3 to test, and crash. DANG. Brought it back down to 4.7, tweaked vcore, ends up crashing. 4.6ghz is okay, but I end up just bringing it down to 4.5 because I think it's better knowing it's stable than having 4.6ghz crash eventually.

I've decided just to leave it at 4.5ghz, which is in the safe zone and ~70-75c at 100% load. Got a hyper212 evo coming in soon but might just leave it alone. Not really sure the performance difference between 4.5 and 4.8 is worth it.


So my final verdict is that I'm thinking that 4.8ghz might be possible with a really high end air heat sink, but I think I would rather leave it at a stable 4.5ghz and prolong its life. Maybe I will try 4.8 again if the hyper212 is more powerful then this sunbeam.

Anyone have luck with 4.8ghz stable on air?

 

[CompuTronix]

Borge80,

Once you reach whatever stable overclock that's achievable at a maximum Vcore of 1.300, you should declare victory and have a beer! I went through five i7 4770K processors before I found one that was stable at 4.7GHz with 1.280 Vcore. It really is the luck of the draw!

You need to read this entire Sticky: Intel Temperature Guide - http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html

" ... Section 8 - Overclocking and Vcore

Overclocked processors can reach up to 150% of their Thermal Design Power (TDP) when using manual Core voltage (Vcore) settings, so high-end air or liquid cooling is critical. Every processor is unique in it's overclocking potential, voltage tolerance and thermal behavior.

Regardless, 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

When tweaking your processor near it's highest overclock, keep in mind that for an increase of 100 MHz, a corresponding increase of approximately 40 to 50 millivolts (0.040 to 0.050) is required. ... "

CT

 

[Borge80]

Nice info CT. Yeah I have settled at 4.5Ghz with a vcore of 1.232. I have attempted 4.7 at 1.292vcore, and 1.280 and unsuccessful. 4.6Ghz is pretty much limbo since I feel it's best to drop it one step past unstable for prolonged life. This chip is still awesome!

Thanks for the help!

 

[CompuTronix]

Enjoy your rig!

 

[killer pc g15]

the die-shrink has nothing to do with the MAX voltage. amd also uses 32nm and they use 1.5volt stock. making a mosfet or transistor smaller does not make it use less volts.
a smaller mosfet / transistor wil use less current then a big one but it can have the same current output! especially the new intel cpus thet use 3dgates.

the gate on a mosfet acts like a capasitor. each time the processor makes a clockcycle it wil empty or fill this capasitor with current till the capasitor reaches the junction voltage this is not the max voltage.
so if you increase your frequency there will flow more current. current x voltage = watts
the junction voltage on most normal transistors is 0.6volts everything higher then 0.6volts whont do anything.
but using a higher switching voltage increases the current flow to the capasitors so that they will reach the junction voltage faster.
http://www.techitoutuk.com/knowledge/electronics/components/capacitors/rccurve.gif
here you can see how a capasitor will charge when voltage is applyed
If you look at the link. and say I use 0.6volt it wil take 15sec to switch the transistor. If I use 1.2 volt I can switch in 7.5sec but if I use 1.8volts its still around 4sec. so you will have to add more and more voltage to get only a little frequency (switching speed) chain.
but what realy kills the transistors / mosfets if the current spikes you are making becouse of adding voltage.

calculating the power: frequency will almost linary increase power consumption. if 3ghz uses 100watts then 4ghz will use 4/3 x 100 = 133watt
when increasing voltage you increase current and voltage at the same time becouse you charge your capasitors to a higher level.
if 1.3volt is 100watts then 1.4volt is (100/1.3^2)1.4^2=115.9 and 1.5volt 133watt

lets say I have a i7 4770k 84watt
3.9ghz and 1.2volt -> overclock -> 4.8ghz and 1.4volt
(84/1.2^2)1.4^2= 114watt -> (114/3.9)4.8= 140watt

 

[4Ryan6]

Idle temps are 28c with sunbeam core contact freezer as a heat sink.

Got a hyper212 evo coming in soon but might just leave it alone.

So my final verdict is that I'm thinking that 4.8ghz might be possible with a really high end air heat sink, but I think I would rather leave it at a stable 4.5ghz and prolong its life. Maybe I will try 4.8 again if the hyper212 is more powerful then this sunbeam.

You're not going to get earth shattering cooling performance results between the Sunbeam Core Contact Freezer (SCCF) and the CMH212EVO, and there's a good possibility you just wasted some money on the EVO.

Both of those coolers are really old school even as popular as the EVO is for it's cheapness, when it comes to raw cooling performance you do get what you pay for.

We all are somewhat concerned with CPU longevity unless you're like the motorcycle guy in the Geico commercial covered with money, and could care less about how long your CPU actually lasts, and I like to see someone that is actually concerned about that.

Controlling CPU load temperature vs applied CPU voltage is very important, and if you are specifically going to be air cooling, the air coolers most high performance capable have to be passive capable coolers, passive meaning can cool to a certain extent with no active cooling fan mounted, just case airflow.

There is a good reason the high end heat pipe air coolers cost as much as they do, they are passive coolers and seriously benefit when cooling fans are mounted to them.

Please take the time and read through this link below so you will be informed of all the air and water cooling possibilities available to you.

http://www.tomshardware.com/forum/id-2196038/air-cooling-water-cooling-things.html

 

[Borge80]

Actually with the evo my temps dropped 10c compared to the sunbeam. Getting 22-25c idle now and 40's under load instead of 35's 50's. My sunbeam was old anyways. CPU is running nice and cool at 4.6ghz.

 

[mikiwroc]

I think, i'm really lucky, because i'm stable with my 4790k at 4,8ghz 1.270 vcore
Today i'll try to get to 5ghz...

 

[HHJ_99_12]

The best i got on air was 4.5GHZ @ 1.12v. Even at that low voltage the temps got high, around upper 70's to low 80's.

I just moved my cpu/mobo to a new case and 360mm water cooler and that resulted in 5.0GHZ stable ! 1.329 volts. I could run 4.9GHZ at 1.26

 

[SinxarKnights]

I think, i'm really lucky, because i'm stable with my 4790k at 4,8ghz 1.270 vcore

Yes mine is like this as well. Glad I got a good sample. 4.8 is my limit before I have to tweak the vcore though . I feel like it can hit 5 but there is no need for me to do so right now.

If you think about it though what is the real difference of 100MHz between 4.4 and 4.5. Seems like a waste of time and electric to do that. If you are only gaming then there is no point. No game even comes close to making the 4790k sweat @ stock. Even if doing CPU intensive task like video rendering, thats only a few seconds saved at best.

 

[Fruityrudy]

Hello my system is stable and temp tested on 4.8 ghz @ 1.25 volt on air cooling
Scythe Mugen max with double sp 140 fans on it
But i want to reach that magic 5.0 ghz so i ordered the Corsair H110i Gtx and that will arrive tomorrow
But what do i need to do with my cache and its voltage? Its at 40 ratio with 1.10 volts manual now with my current overclock
If i go for the 5.0 ghz can i leave those values alone for stability ?

I7 4790k

 

[HHJ_99_12]

Hello my system is stable and temp tested on 4.8 ghz @ 1.25 volt on air cooling
Scythe Mugen max with double sp 140 fans on it
But i want to reach that magic 5.0 ghz so i ordered the Corsair H110i Gtx and that will arrive tomorrow
But what do i need to do with my cache and its voltage? Its at 40 ratio with 1.10 volts manual now with my current overclock
If i go for the 5.0 ghz can i leave those values alone for stability ?

I7 4790k

I didnt mess with any of the cache settings. I got my 4790K to 5GHZ @ 1.33v