Help me i7 4790k very hot !!!

[MrFuny]

--20.12.2015--------------------------------------------------------
Thanks for all my friends,
Now I setting 1.2 volt for turbo 4.4G and 1.28volt for 4.6G - My Problem may be solved

Hi All,

I just new PC :

CPU: i7 4790K
Mobo: GA-Z97M-D3H(new mobo GA-Z97X-UD3H)
RAM: 2 x Mushkin Enhanced Blackline Pro 8GB ( 1x8GB ) bus 2133 cas 11 - DDR3
cooler: Deepcool Gamer Storm Maelstrom 240 - AIO Liquid Cooler

-with my setting( all default, only set turbo for all cores up to 4.4G - default 4.2g), power consumption show up 184W normal or problem?

 

[buildzoid]

79C is about what I would expect with those settings. Maybe even more.

 

[rtwjunkie]

Are those volts normal for Devil's Canyon with basically a small overclock? I'm only asking because I don't know, and 1.33 looks a little scary/heat-producing to me.

 

[EarthDog]

That is in the ballpark, sure.

~1.4v would be my daily driver limit assuming temps are in order.

 

[GhostRyder]

For that cooler, that seems a bit high as on hyper212 Evo my friends 4690K (Yes an i5 but they are similar enough) is overclocked to 4.5ghz and keep in the low 60's under prime95.

Are those volts normal for Devil's Canyon with basically a small overclock? I'm only asking because I don't know, and 1.33 looks a little scary/heat-producing to me.

I agree, but its not that high and should not cause those temps.

 

[EarthDog]

This is IBT, not P95. Can't compare apples to oranges as they are different applications. Each chip can be a different temperature even at the same voltage anyway (leakage).

 

[GhostRyder]

This is IBT, not P95. Can't compare apples to oranges as they are different applications. Each chip can be a different temperature even at the same voltage anyway (leakage).

I know, even so both are intended to push chips very hard and the problem here is the temps for 4.5ghz on a hyper212 which is a cheap cooler are outdoing a 240 AIO. He should be getting lower temps even using (Edit) IBT.

I didn't have an exact reference, but I still believe something is going on for just 4.4ghz.

 

[trog100]

"(Yes an i5 but they are similar enough) "

no they are not.. at least not when cpu burn in software that uses the hyper threading is being run.. one runs 4 threads the other 8 threads.. one hell of difference temperature wise..

having said that.. in normal use they are very similar.. just not when all 8 threads are being run balls out.. i recon a 10 to 15 C tempt difference will show up..

1.33 vcore is a tad (a lot to be honest) on the high side for 4.4 gig.. my 4790 will run a stable 4.4 at 1.2 vcore.. the first thing to do is lower the vcore until the cpu bombs.. then go up a tad..

use something like wprime for heat testing.. the intel burn stuff is very cruel.. it will run f-cking hot and its meant to..

trog

 

[peche]

no they are not.. at least not when cpu burn in software that uses the hyper threading is being run.. one runs 4 threads the other 8 threads.. one hell of difference temperature wise..

the most important fact by the way .... i5's are excellent because the relation between their performance and temperatures... most of my friends have i5's compared to my i7 when was stock [not delidded or OC'ed] temps were pretty far distance talking about temperatures... when i decided to delid ... it shown some abysmal improvement on temps... but in some chases like a delidded i5 3570k of one of my friends... temps still some high on my chip ... difference is between 10C on idlle and full load...both running the same cooler, [Tt's water 3.0 pro ...we have the same cooler! ]

 

[EarthDog]

use something like wprime for heat testing..

No. WPrime test is WAY too short for temperature testing... It takes an air cooler at least 15 minutes to get to a saturation point, water, even longer.

 

[GhostRyder]

"(Yes an i5 but they are similar enough) "

no they are not.. at least not when cpu burn in software that uses the hyper threading is being run.. one runs 4 threads the other 8 threads.. one hell of difference temperature wise..

having said that.. in normal use they are very similar.. just not when all 8 threads are being run balls out.. i recon a 10 to 15 C tempt difference will show up..

1.33 vcore is a tad (a lot to be honest) on the high side for 4.4 gig.. my 4790 will run a stable 4.4 at 1.2 vcore.. the first thing to do is lower the vcore until the cpu bombs.. then go up a tad..

use something like wprime for heat testing.. the intel burn stuff is very cruel.. it will run f-cking hot and its meant to..

trog

Uhh no, it may make the processor heat up a slight bit more because its keeping it at a higher usage more efficiently but its not night and day in the differences temp wise. If I turn off Hyper-Threading on my 5930K and run burn in tests the temp difference is negligible because its already stressing the cores to their limits. 4 core are 4 cores, only in tests where the cores are not being used efficiently (Or programs) is where there will be a temp difference as the Hyper-Threading is there to help when cores are not being used to their fullest so there is no waiting around.

Not to mention in this case he's at 79c on a 240mm AIO versus a cheap air cooler keeping temps down in the 60's in similar burn in tests at higher clocks. This is still no matter how you view it unusually high for a devils canyon especially on that cooler.

 

[EarthDog]

I see around 10C difference in temps with HT enabled vs without. I have seen this behavior since Nehalem through Ivybridge-E.

 

[vega22]

op ignore the bitching if you can, your temps look ok to me for that voltage.

i do not think you need that much vcore for that clock speed.

 

[GhostRyder]

View attachment 70120 Hi All,

I just new PC :

CPU: i7 4790K
Mobo: GA-Z97M-D3H
RAM: 2 x Mushkin Enhanced Blackline Pro 8GB ( 1x8GB ) bus 2133 cas 11 - DDR3
cooler: Deepcool Gamer Storm Maelstrom 240 - AIO Liquid Cooler

I use IBT 2.54 burn CPU with turbo All cores 4.4Ghz(others default).

But I think it very hot and power consumption not good! Please check attack image.

OP, did you set those clocks or are you referring to the stock settings of the chip (Base 4.0ghz with Turbo to 4.4ghz)?

 

[hat]

It seems we're having a bit of a war here on whether HT increases temps or not (or, by how much). I'll have to vote that HT does have a rather significant impact on temps, as my experience with an i7 920 and various forms of stress testing always indicated that HT on meant more heat.

 

[trog100]

i thought about disabling HT on my 4790K to drop the tempts whilst running daft things like prime95.. but then again i also noticed that in real world gaming and the like disabling HT made no difference.. so i left it running..

to be honest a cpu soon gets within 2 C or so of its maximum temps when air cooled.. two or three minutes is plenty.. it soon drops back down as well..

any significant heat after that has bugger all to do with the cpu.. its all down to other stuff slowly warming up.. my graphics cards for example take a good five or ten minutes to slowly ramp up to maximum tempts.. my cpu less than two minutes..

my favourite heater for both graphics cards and cpu is furmark.. it has its own cpu burner as well as the gpu burner.. it runs maybe one or two C hotter than wprime does..

wprime on its on its longer test makes a good quick tempt check.. it can also be run several times if one run isnt enough.. but i like furmark.. and furmark is what i use..

i simply leave the furmark burner running whilst browsing and doing other stuff for cpu heat and basic stability checks .. to test total system heat i run the heaven benchmark and the furmark cpu burner at the same time.. i would run the furmark gpu test at the same time as its cpu burner but the furmark gpu test needs a little more cpu grunt to run than heaven which in reality needs bugger all cpu grunt..

i am sure the ops vecore is too high.. these auto clocking things tend to do that if that is what is being used.. i tried the gigabyte stuff that came with my board.. for what it called a mild 4.5 gig overclock it ramped the vcore up to 1.35 way too high.. my system is now running 4.5 gig at 1.22 vcore its also showing 24 C while i type this.. not as i think idling temps mean much..

i think the max vcore needed for a 4790K at 4.4 would be 1.25.. but one thing is for sure.. with all 8 threads being run balls out a 4790K aint a cool chip.. 4.4 is just cruising for it though.. which is way it dosnt need much vcore voltage..

hyper threading does cause more heat to be generated.. the chip does more work and logically gets hotter.. those who think it dosnt are wrong.. clock for clock an I7 will run a fair bit hotter tan an I5 when the software is running 8 thread on 4 cores.. if it didnt there would be no point in having one.. he he

trog

 

[Toothless]

Those temps are fine, same with the volts. Why did the thread have to break out into war?

 

[EarthDog]

Trog... Oy...

1. Disabling HT while running a stress tests defeats the purpose of the stress test!!! You can be stable with HT off but when you enable it, it's unstable. You can actually use a bit less voltage with HT disabled
2. In order to empirically compare, you need to use the same exact chip with HT. Test with it on, disable, test with off. Comparing two different chips wouldn't prove anything.
3. While his VT age is likely high for the clocks and auto does tend to overvolt in favor of stability, your 'rules' about voltage and clocks are not black and white as you appear to be saying.

@Toothless - "war" (I say discussion, i didnt see anyone out of line) is to correct misinformation. It comes from a few members here in piles..

Oh.. "hehe"

 

[Jetster]

Dont use Intel burn test

 

[MrFuny]

This is IBT, not P95. Can't compare apples to oranges as they are different applications. Each chip can be a different temperature even at the same voltage anyway (leakage).

But its look like scary: CPU package ~184W, I think not good for GA-Z97M-D3H.
Also, I not sure 184W is normal for 4790K and my setting(All default, only turbo 4.4G for all cores), when burn CPU with IBT - My CPU have problems?

Dont use Intel burn test

Why?
// I'm nood

"(Yes an i5 but they are similar enough) "

no they are not.. at least not when cpu burn in software that uses the hyper threading is being run.. one runs 4 threads the other 8 threads.. one hell of difference temperature wise..

having said that.. in normal use they are very similar.. just not when all 8 threads are being run balls out.. i recon a 10 to 15 C tempt difference will show up..

1.33 vcore is a tad (a lot to be honest) on the high side for 4.4 gig.. my 4790 will run a stable 4.4 at 1.2 vcore.. the first thing to do is lower the vcore until the cpu bombs.. then go up a tad..

use something like wprime for heat testing.. the intel burn stuff is very cruel.. it will run f-cking hot and its meant to..

trog

thanks, I will setting again my System, with lower volt.
But, I'm not sure 184W is nomarl for i7 4790k - with my setting.

Those temps are fine, same with the volts. Why did the thread have to break out into war?

temps are fine?
Power consumption?
// I'm nood pls help me

OP, did you set those clocks or are you referring to the stock settings of the chip (Base 4.0ghz with Turbo to 4.4ghz)?

Default turbo for all cores only 4.2G. But I set turbo for all cores up to 4.4G.
When I burn CPU with IBT, Hardware monitor show up: Power consumption of CPU Package is 184W and temp 79*C. I don't know this is normal for i7 4790k

 

[rtwjunkie]

@MrFuny just to help you out and save you some grief later from moderators, please try not to double, triple and quad post.

Just use the multiquote button and the insert the quotes into your response area (there's a button for that too at the time).

 

[trog100]

the power consumption (usage) should be directly related to your clock speed and your vcore.. your vcore is higher than it needs to be for 4.4 thats for sure..

4.4 for a 4790 chip isnt over working it.. i cant say about the 184 watts.. i have never measured mine..

but the first thing to do is lower the vcore and report back..

as a guide my own chip runs 4.2 stable at 1.117 vcore.. 4.4 stable at 1.2 vcore.. 4.5 stable at 1.22 vcore.. after that the voltage needs to go up.. at 4.6 i need a vcore of 1.28..

i would need at least 1.35 for 4.7 but i have never really bothered going there.. things get too hot and i know i am going beyond the chips "happy" speed.. .. he he

pick a clock speed you want to run at.. for general usage 4.4 is fine.. set a vcore of 1.25.. if things are stable and think they will be start going lower until the chip bombs.. when it bombs take it up a tad..

it isnt just the vcore that affects the temps.. when the chip runs faster it does more work and generates more heat.. the intel burn in test is more of a survival test than anything else.. it should not be used as a real life indication of anything other than the chip (system) will run it without failing..

trog

 

[Divide Overflow]

Is that the lowest stable vcore you can get for those clocks?
I'd try getting it down some. It's safe to run 1.35 but you can probably do better.

 

[MrFuny]

the power consumption (usage) should be directly related to your clock speed and your vcore.. your vcore is higher than it needs to be for 4.4 thats for sure..

4.4 for a 4790 chip isnt over working it.. i cant say about the 184 watts.. i have never measured mine..

but the first thing to do is lower the vcore and report back..

as a guide my own chip runs 4.2 stable at 1.117 vcore.. 4.4 stable at 1.2 vcore.. 4.5 stable at 1.22 vcore.. after that the voltage needs to go up.. at 4.6 i need a vcore of 1.28..

i would need at least 1.35 for 4.7 but i have never really bothered going there.. things get too hot and i know i am going beyond the chips "happy" speed.. .. he he

pick a clock speed you want to run at.. for general usage 4.4 is fine.. set a vcore of 1.25.. if things are stable and think they will be start going lower until the chip bombs.. when it bombs take it up a tad..

it isnt just the vcore that affects the temps.. when the chip runs faster it does more work and generates more heat.. the intel burn in test is more of a survival test than anything else.. it should not be used as a real life indication of anything other than the chip (system) will run it without failing..

trog

 

[Jetster]

Why?
// I'm nood

Because the Intel Burn Test is excessively harsh, unnecessary and will reach 10c higher temps than any other stress test. if you have bad surface contact with the cooler you could very quickly put the chip in danger. There are many other stress test that are more appropriate for testing your system.

OCCT, AIDA64 are some of my favorite