• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Question regarding 4790K OC'ing using the Intel Extreme Tuning Utility

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
ilikefinefood, in one of your early posts you state you don't know anything about the memory. If you will download and install CPU-z and click on the "Memory" tab and the "SPD" tab it will tell you about your RAM. Please do that and post pics of those two tabs. If you are using higher speed memory then that could explain the System Agent offset.

In fact, the time has come for you to create a Signature (or "Sig" as we call it) so that we can have snapshot of your system components that travels with every post you make.

Sure, here you go mate

memory.png

- - - Updated - - -

No probs.... shall lower to 0.05
 
Is VIN4 supposed to be lowering with VID ? whatever VIN4 is ?
I'm going to be frank... don't know and don't care (for this thread. :)). Who knows what it is, but it has no bearing on helping this guy out... I do know that. :)

There are already enough hands in the cookie jar here... I don't play these adaptive/offset voltage games, its manual and donezo. :)
 
You can crop the white space too, LOL! You;ll get it! :)

You still peaked at 1.369V. Try .05 offset so you peak around 1.3v.


Regardless though, those temps are way too hot for your cooler and these settings. Something isn't right there. Are you sure the pump is on and circulating? Im mean, your LOW temps are 75C which is way WAY too high. YOu should idle around 30C give or take a couple C.

Ill get there in the end with cropping lol


Ok, they should be helpful to you, id go back and check em out. :)

Your total vcore is still very high. Lower your offset to .05 to get those temps down.

shall lower to 0.05 and try again
 
Sorry, ED. Missed the fact that you already suggested .05 offset in your earlier post. Lots of rapid fire posting going on in this thread.

Offline for a bit now.
 
I think he has a cooler problem personally trents... Id almost want to start over and see if his temps are bad at stock...

@ ilike - Look at his low temps (though that may have been an after the test type thing). typically we ask the program to be open on idle, then go into the stress testing app. This way your true idle temps are caught and max high temp.
 
Ok, with a .05 offset the test started off looking good, no higher than 85 degrees..... As the first phase ended on prime, everything passed the temps shot up to what you see below :(

fhfhfhfh.png
 
Ok, idle temps here are good... so the cooler is likely doing its job. P95 is likely hitting those FMA3 instructions and jacking the voltage up (another reason I don't like offset/adaptive...........). Try running AIDA64 Stress test instead. P95 and those FMA3 instructions sets are killers and rarely used by most anyway.
 
I think he has a cooler problem personally trents... Id almost want to start over and see if his temps are bad at stock...

@ ilike - Look at his low temps (though that may have been an after the test type thing). typically we ask the program to be open on idle, then go into the stress testing app. This way your true idle temps are caught and max high temp.

This is while idling on the desktop @ 44 - 0.05

idle.png

- - - Updated - - -

Ok, idle temps here are good... so the cooler is likely doing its job. P95 is likely hitting those FMA3 instructions and jacking the voltage up (another reason I don't like offset/adaptive...........). Try running AIDA64 Stress test instead. P95 and those FMA3 instructions sets are killers and rarely used by most anyway.
 
It didn't like that AIDA64..

Started off slugish , looked like temps were at roughly 70-75 though on the bottom window the CPU usage was obviously at 100% it said 'overheat detected' in red writing ..... The program self terminated shortly after :(
 
Ok, after 7 minutes of testing using this older version here are my findings - temps a lot lower. No errors or warnings across all cores

hhhhhhhhhhhhhh.png

Even got the hang of posting pics lol - things MIGHT be looking up :/
 
You are on fire with the pics!!! Certainly appreciated! :attn:

What I take away from that is this....

1. It wasn't FMA jacking up your voltage...
2. Temps still seem warm to me for that cooler and these clocks.


To that end:

A. Unless you know what your starting load voltage is, I would start over. Reset all voltages for the CPU to auto and reboot. Load P95 again, note the max/running voltage (post here). This is your starting/stock voltage. When using offset you are adding to that voltage. Here is the thing, that offset is giving the CPU too much voltage for whatever reason. For 4.4Ghz, I can't imagine 1.3V is needed. 1.2-1.25V is what I would expect there. I would see if you can get 1.25V load voltage and see how that works out for you as far as temperatures and stability. To be frank again..... I would scrap this offset garbage and just run manual as its easier, LOL!
B. What is your room temperature? How are the fans configured in your case (where and what direction are they all blowing?)?
 
I think he has a cooler problem personally trents... Id almost want to start over and see if his temps are bad at stock...

@ ilike - Look at his low temps (though that may have been an after the test type thing). typically we ask the program to be open on idle, then go into the stress testing app. This way your true idle temps are caught and max high temp.

I agree.

ilikefinefood, we are concerned that your water cooler is not working correctly.

So go into bios and hit the f5 key to restore to system defaults with no overclock, no extra voltage. Then run Prime95 for 20 minutes. At about the 18-20 minute mark temps will start to ramp up. That is normal. We need to see what your temps are like at stock speeds and voltages. Also, have HWMonitor already open before you start the stress test so we can see what idle temps are like.
 
Trents... read the entire thread (from where you left off), then post... you are woefully behind again as that post was like 45 mins ago. :)
 
You are on fire with the pics!!! Certainly appreciated! :attn:

What I take away from that is this....

1. It wasn't FMA jacking up your voltage...
2. Temps still seem warm to me for that cooler and these clocks.


To that end:

A. Unless you know what your starting load voltage is, I would start over. Reset all voltages for the CPU to auto and reboot. Load P95 again, note the max/running voltage (post here). This is your starting/stock voltage. When using offset you are adding to that voltage. Here is the thing, that offset is giving the CPU too much voltage for whatever reason. For 4.4Ghz, I can't imagine 1.3V is needed. 1.2-1.25V is what I would expect there. I would see if you can get 1.25V load voltage and see how that works out for you as far as temperatures and stability. To be frank again..... I would scrap this offset garbage and just run manual as its easier, LOL!
B. What is your room temperature? How are the fans configured in your case (where and what direction are they all blowing?)?

No idea what my starting voltage was but one thing for sure is I haven't changed ANYTHING other than the offset and the core multiplier......... When I started off here today I pressed F5 in the bios and started from there with all of the advise/recommendations in this thread.

So you're saying I should F5 it again in the bios and ONLY change the multiplier to 44 - then boot and run the test again?

thanks as always
 
Sorry I wasn't clear. F5 again, but leave the CPU multiplier at auto as well. We want stock clocks, not what your PC thinks is good for 44x multi which is what would happen if things are on auto except the multi.


Again, what is your room temp? How are your fans configured and what direction are they blowing?
 
I agree.

ilikefinefood, we are concerned that your water cooler is not working correctly.

So go into bios and hit the f5 key to restore to system defaults with no overclock, no extra voltage. Then run Prime95 for 20 minutes. At about the 18-20 minute mark temps will start to ramp up. That is normal. We need to see what your temps are like at stock speeds and voltages. Also, have HWMonitor already open before you start the stress test so we can see what idle temps are like.

Trents... read the entire thread (from where you left off), then post... you are woefully behind again as that post was like 45 mins ago. :)

Sorry I wasn't clear. F5 again, but leave the CPU multiplier at auto as well. We want stock clocks, not what your PC thinks is good for 44x multi which is what would happen if things are on auto except the multi.


Again, what is your room temp? How are your fans configured and what direction are they blowing?

So as Trents said above, F5 totally stock and run prime95 (new version) again for 20 mins?

the room is quite cool TBH, im in england and the radiator/central heating is turned off at the radiator to purposefully keep the room cool.

As for the fan's, the Corsair cooler is front mounted, I have a fan on the side grill above the GPU and CPU sucking in and one on the back sucking out
 
F5 totally stock and run either or, or both P95s...

So, you do not have any fans on the radiator?????????????????? Or are there two fans there, both blowing in the case I would guess?

What case do you have? Can you please take a moment to full out your signature (see trents request earlier) so we know exactly what you are working with? :)
 
Back