• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

***Intel i7 4790K Owners thread***

Skiddley said:
Interesting. Might give the vixed voltage a go. I was less concerned about the cost of the additional power running a fixed vCore, and more about the impact to idle temps. I see vCores drop to ~0.8 at 800mhz you see. How big is the impact, Minimal?

How do you disable the turbo in an ASUS UEFI BIOS, can't seem to see the option. Is it done by default whan you sync all cores?
Click to expand...

I'm not sure how it's disabled on ASUS BIOS, here's where it's changed in mine. The screen cap is from my 4670k @ 4.4 but it's the same idea for 4790k.

X2xWtSC.jpg


Vcore was 1.22...


BiYHFSy.jpg



And LLC was set to extreme (note the description on the right hand side, explaining the effect on voltage)...



eDLlODA.jpg
 
Skiddley said:
Can't extreme LLC settings boost vCore above the configured value under load?
Click to expand...

No, the opposite is true. If I dial it down a notch, the vcore will fluctuate a little. The lowest setting has it fluctuating all over the place.

When set to extreme, the vcore never changes regardless of core clock.

Experiment a little with yours. Set a mild overclock and see what effect the different LLC settings have when under load. Just be careful, because when I set mine at the opposite end of extreme the vcore can fluctuate a lot.
 
On the last 2 GA boards i have had LLC made no differance what ever the setting the vcore was rock solid. However on other board yes the vcore can over volt a little.

With my 4670k @ 4.2Ghz with 1.278v under full load tests i got a 0.002 drop when the cores was first loaded up then it was just rock solid at the set vcore with LLC on high.
I will be tested with a 4690k today.
 
Little reluctant to alter things as it seems to be happy, and temps are good. @4.7G on a +0.030 adaptive now and stable. Firestrike is giving >13K with a 980.

I think I'm set for GTA :D
 
So I won the thermal lottery after all?
I oc'ed the chip finally. But I'm in disbelief wheter its real or I'm just dreaming.

I have dynamic mode on and multiplier into all cores rather then per core.
I set vcore to 1200 just for safe starter point to increase multiplier. After every +1 multiplier I played some FC4 and DAI to see it in action a little, whilst keeping HWmonitor and CPU-Z active in background and monitoring if everything is ok.

After seeing that 4.7Ghz on 1.2v seemed ok, I changed vcore to 1.225 just to be sure and set x48 multiplier. Perhaps I should try 1.2 again and see how much I can lower it?

Since I didnt need to increase vcore much at all, tempratures are really really good. We're talking 18-24c idle and 43-48c load on Dragon Age Inquisition and Far Cry 4 I have tried so far.
This is all air cooled with a Zalman CNPS 9900 MAX, a Corsair Obsidian 450D filled with 7x Corsair AF Quiet edition 120mm.
Fans were all running at 100% and cpu fan around 70-80%.

Up until now, I've had 0 BSOD or any weird stutters/artifacts whatever...so how do you guys recommend me to continue from here? 5Ghz is not possible on these chips right?

Update: So I just tried 4.9 GHz at 1.225v, and had my first freeze on DAI. I will try 1.250v and cpu fan on 100%.
So 4.9Ghz at 1.250 seems fine so far. Tempratures are rising though. Far Cry 4 at 45c to 50c, Dragon Age Inquisition even higher at 48-55c average 51-52c. So yeh I'm looking if I should just try 5Ghz.

So I tried 5GHz at 1.295v, playing some Battlefield 4 Shanghai campaign level. Tempratures varied around 55-58c. Unfortunately after playing for a while, game froze eventually. Though for some reason I can't push more then 1.3 vcore, seems there is a safety tresshold for anything above 1.3v. I use MSI Z97 gaming 5 mobo. I guess 1.3v isn't safe anyhow or is it?
 
Last edited:
String said:
No, the opposite is true. If I dial it down a notch, the vcore will fluctuate a little. The lowest setting has it fluctuating all over the place.

When set to extreme, the vcore never changes regardless of core clock.

Experiment a little with yours. Set a mild overclock and see what effect the different LLC settings have when under load. Just be careful, because when I set mine at the opposite end of extreme the vcore can fluctuate a lot.
Click to expand...

Is it fluctuating under load or fluctuating with varying load due to power saving tech? How are you measuring the voltage?
 
PullTheTricker said:
So I won the thermal lottery after all?
I oc'ed the chip finally. But I'm in disbelief wheter its real or I'm just dreaming.

I have dynamic mode on and multiplier into all cores rather then per core.
I set vcore to 1200 just for safe starter point to increase multiplier. After every +1 multiplier I played some FC4 and DAI to see it in action a little, whilst keeping HWmonitor and CPU-Z active in background and monitoring if everything is ok.

After seeing that 4.7Ghz on 1.2v seemed ok, I changed vcore to 1.225 just to be sure and set x48 multiplier. Perhaps I should try 1.2 again and see how much I can lower it?

Since I didnt need to increase vcore much at all, tempratures are really really good. We're talking 18-24c idle and 43-48c load on Dragon Age Inquisition and Far Cry 4 I have tried so far.
This is all air cooled with a Zalman CNPS 9900 MAX, a Corsair Obsidian 450D filled with 7x Corsair AF Quiet edition 120mm.
Fans were all running at 100% and cpu fan around 70-80%.

Up until now, I've had 0 BSOD or any weird stutters/artifacts whatever...so how do you guys recommend me to continue from here? 5Ghz is not possible on these chips right?

Update: So I just tried 4.9 GHz at 1.225v, and had my first freeze on DAI. I will try 1.250v and cpu fan on 100%.
So 4.9Ghz at 1.250 seems fine so far. Tempratures are rising though. Far Cry 4 at 45c to 50c, Dragon Age Inquisition even higher at 48-55c average 51-52c. So yeh I'm looking if I should just try 5Ghz.

So I tried 5GHz at 1.295v, playing some Battlefield 4 Shanghai campaign level. Tempratures varied around 55-58c. Unfortunately after playing for a while, game froze eventually. Though for some reason I can't push more then 1.3 vcore, seems there is a safety tresshold for anything above 1.3v. I use MSI Z97 gaming 5 mobo. I guess 1.3v isn't safe anyhow or is it?
Click to expand...

You need to test with IBT / Realbench and other stress tests to see the true stability of this chip. My 4790K can run games at 4.8GHz but bluescreens when running Prime95 for 10-20 mins. Even 4.7GHz takes 1.3V for stability so it appears to be a middle of the road cpu.

You could have a golden chip but try running Prime95 version 27 or older and see if it is stable for a couple of hours.
 
fs123 said:
You need to test with IBT / Realbench and other stress tests to see the true stability of this chip. My 4790K can run games at 4.8GHz but bluescreens when running Prime95 for 10-20 mins. Even 4.7GHz takes 1.3V for stability so it appears to be a middle of the road cpu.

You could have a golden chip but try running Prime95 version 27 or older and see if it is stable for a couple of hours.
Click to expand...
Ye I was planning to, but I'l have to do it with caution. Such programs run your hardware into situations that are unrealistic in daily usage, so not sure how safe it is.
 
Stress tested AIDA64 and Realbench at 4.8Ghz at 1.225v. I immediately raised vcore from 1.200 to 1.225. Is a 0.025 jump too large? Not sure how you guys go about it.
Eitherway, so far it seems stable I guess and core tempratures from 65-70c. Thats on the high side, but its normal for a stress test right? If so, would this give a green light to keep the system at 4.8Ghz 1.225v 24/7.

I checked Battlefield 4 to see for any perfomance gains. Couldn't see none, but I only paid attention to max fps, rather then minimum fps wich is actually most important. I wonder though, is it worth it?
 
PullTheTricker said:
Stress tested AIDA64 and Realbench at 4.8Ghz at 1.225v. I immediately raised vcore from 1.200 to 1.225. Is a 0.025 jump too large? Not sure how you guys go about it.
Eitherway, so far it seems stable I guess and core tempratures from 65-70c. Thats on the high side, but its normal for a stress test right? If so, would this give a green light to keep the system at 4.8Ghz 1.225v 24/7.

I checked Battlefield 4 to see for any perfomance gains. Couldn't see none, but I only paid attention to max fps, rather then minimum fps wich is actually most important. I wonder though, is it worth it?
Click to expand...


You're safe all the way up to 1.35v on Air for 24/7. 4.8Ghz at 1.225v is excellent. I use 1.4v+ on 4670k for 4.8Ghz+ when running benches. Oh and 85c-90c is fine but you probably wouldnt hit that gaming, only when running AIDA64 etc. I'd change Vcore to 1.25v and try 4.9Ghz :D

I don't believe in running Prime95 and others to see if its stable, just use the PC as you would normally . gaming etc.
 
Dempsey22 said:
You're safe all the way up to 1.35v on Air for 24/7. 4.8Ghz at 1.225v is excellent. I use 1.4v+ on 4670k for 4.8Ghz+ when running benches. Oh and 85c-90c is fine but you probably wouldnt hit that gaming, only when running AIDA64 etc. I'd change Vcore to 1.25v and try 4.9Ghz :D

I don't believe in running Prime95 and others to see if its stable, just use the PC as you would normally . gaming etc.
Click to expand...
My bios doesn't allow me to go further then 1.3v, or otherwise I could probably get 5Ghz stable at perhaps 1.315v I would imagine. But you know, its a trade off, I want my cpu to have longevity and not wear down quickly.
But I think its obvious my sample is a silicon lottery being able to do this on air.
 
Yeah i would certainly try and get more out of the chip, its proberbly only gonna take another 0.05v to get 4.9 which will still leave ya way under 1.3v
 
BluD said:
Yeah i would certainly try and get more out of the chip, its proberbly only gonna take another 0.05v to get 4.9 which will still leave ya way under 1.3v
Click to expand...
4.9Ghz went pretty easily and stable with 1.245v. Its just the 5Ghz that requires probably around 1.305v or around that. Googled it up why bios doesnt allow me to go beyond 1.3v, no luck though. I wonder if its a safety tresshold set by the manufacturer.
 
So I finally troughly tested the 5Ghz clock at 1.350v. Ran some Far Cry 4 for an hour, tempratures okish at 50-60c maximum. It never raises above 60 so thats good I guess.
But then, the tests on Aida64 and Realbench worry me. Tempratures on these stress tests were going anywhere from 78c to 85c and 90c at peak.
Thats insane, and makes me wonder if a 5GHz OC is worth it when I'm running on air. How much temprature could I reduce by upgrading from a Zalman CNPS 9900 MAX, to a high end water cooler? If its close to 10c then I can see myself running 5GHz 24/7 I guess... if not then its probably not worth it.

Seems like 4.8 Ghz at 1.215v - 1.225v is best ''bang for buck''.
 
GiraffePencils said:
your results on air are making me think your chip is magic
Click to expand...
Yes I am surprised as well. Before this I had a Q6600 B3 stepping and it downright sucked, worst chip ever. I'm not an enthousiast oc'er in any way, but I guess this time I was lucky anyway.

Well infact the 5Ghz on 1.295v I played Battlefield 4 for quite a while before the pc froze... thats a testament in itself that I prolly got a rare chip here. Wich makes me reconsider on water cooling. But I allready paid 50 euro's for my air cooler, but I couldn't have predicted this at all, quite surprised me actually.
What would you guys do in my situation?
 
You must log in or register to reply here.
Back
Top Bottom