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

Lets discuss 105º on Ivy

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Not sure if I explained myself properly...(or did it any better below) but that doesnt make sense at all to me. Doesnt matter how it gets there, its still going to get there, and QUICKLY. It makes little difference. Now, if you are NOT hitting full boost, I agree. But normally you hit full boost doing anything strenuous on the rig.
 
Not sure if I explained myself properly...(or did it any better below) but that doesnt make sense at all to me. Doesnt matter how it gets there, its still going to get there, and QUICKLY. It makes little difference.

The end game is the same.. so upon load temps will be the same regardless if the voltage steps up to get there or is there. If you sit at 5Ghz with 1.4v idle temps will be higher. When you boost to 5Ghz and step to 1.4v, the load temps will be the same shortly after. I see what you are saying but, the time it takes to reach the same temperature is nearly nothing from say 40C idle and 1.4v at 5Ghz to 30C idle and stepping up to 5Ghz.

You got a big part of the picture there but the one thing about the offset is I have seen many times where using the offset will drop the voltage even at max OC and under load. Less heat.
 
What cooler do you have? try to put half the cpu's cashe an fft and half the ram to see if your system reboots
 
Interesting. I do not use offset, but have not seen that reported behavior before. I think there is a thread here on offsets and SB where that was speculated (less voltage for an overclock) but I think it was anecdotal.

Regardless. Thanks for the info. :)
 
Well I gave up on setting the offset manually. I always get funky results. I also never take LLC over medium. The reason for the LLC med or less is the dynamic vcore compensates as well raising volts even more (in theory) creating more heat. I use + offset in Auto.

EDIT: I am enjoying this thread because it is what the spirit of OCF is about. Discussion not argument and collective learning to help speed up our systems..
 
To find my cpu lowest operating voltage i'm doing all the cache to make sure is fully powered the ram is at half t as i loose 1 core very fast with an error this way.

Stabilities tests i run for 2 hours why i have the impression this cpu will reboot?
 
Manual and LLC here.. enough LLC to have juuuuuuuust a tad of vdroop.

Well the fact that you are not scared of droop is good some people take it to far and ruin chances at a high OC. Better a great PSU, excellent MB PWM than Cheap PSU and PWM with high LLC.
 
LLC is awsome. It's the foundation of an awsome oc :thump: I hear it's on AMD boards too. Vdroop is the reason why I lost so many AMD boards when I ran em, I think. Heard the zzzt a few times with em lol
 
PSU and CPU vdroop are two entirely different operations.
The PSU droops too, but that is entirely dialed out by the CPU buck converters.

A good PSU is definitely an excellent idea, but it won't have more or less CPU vdroop than a cheaper unit.
(Like any blanket statement there are exceptions, if your PSU is putting out 9v on the 12v rail everything will suffer. I wouldn't be surprised if vcore stayed right where it was though)

For my daily OC I like offset vcore for the low idle voltage.
For benching, I like/need/want significantly over stock at idle so I usually use enough static to keep that happy, and then offset/dynamic for the load vcore if the board demands it.
If the board doesn't require offset though, I'm all static for benching.
Benching, of course, is a different story anyway.
 
I guess it was a bit of a blanket statement. I understand what the two are, and I see them in action everyday. It reminds me that my current psu has seen better days :D
 
Here is my view on it and why I even mentioned the PSU (And yes PSU power is not the same as CPU power but it can have an affect; though, more-so the power going to the board than to the CPU.

If I buy a cheap board and PSU I will most likely not have stable power from the PSU or to the CPU. So in attempts to rectify these issues I may try LLC and though it may give me 100Mhz it will do more to create heat than anything else forcing me to down-clock my CPU back to the previous level.

I am also not saying that LLC is bad I am just saying many people use it and do not need it set at its highest level. Better to have response times set at their fastest.
 
ambient temrature is missing

Are we talking Intel Tech Doc ambient here, Case ambient or room ambient?

If it is the later two it is a moot point because the effect that is seen on each individuals setup is dependent on Case ventilation, CPU cooler and of their own personal hose temperature preferences.
 
Last edited:
Just wanted to say thanks for the thread. I am learning all the time from you guys. At times you talk over my head but regardless I am still reading and learning. I hope to have all my parts by September and tackling my first build an OC build at that.
 
I just wanted to share my findings with this chip. I didn't get a gem by any means, setup required 1.36 vcore to run 4.5ghz. Temps were no higher than 82C running Prime 95. I just left it like that for the last few weeks and then all of a sudden I started getting BSOD's when gaming. I knew I was pushing it with that overclock, but it was prime stable so I left it that way. I went from prime stable at 4.5 to not even close to prime stable at 4.5 over a few weeks. I am convinced that the high voltage degraded my chip ever so slightly. I turned it down to 4.2 @ 1.2 and it runs without a hiccup now.
 
Got your PM, but I am afraid I cannot help much here. That Tom's dude, bubba-hotepp or someshhng like that started a thread about AMD's issue with TCase, TJmax etc that was incredibly informative, but not sure if that would translate to this at all.

Subscribed though...:)

Last I heard (and I'm not 100% certain as I haven't fully researched Intel CPU's) TJunction was the temperature to watch with an Intel CPU.
 
the one thing about the offset is I have seen many times where using the offset will drop the voltage even at max OC and under load. Less heat.

You mean offset allows for a lower load voltage (and heat) for a given overclock then fixed?
 
I tried offset exactly 1 time. Set cpu to default, set voltage to default. Rebooted, put offset at .005 to see what reading I would get. Loaded prime95, watched cpu-z report 1.65 volts, up from lile 1.2 at idle. Hit the power switch, went back into bios and went back to manual.
 
Last I heard (and I'm not 100% certain as I haven't fully researched Intel CPU's) TJunction was the temperature to watch with an Intel CPU.

tjunction is for mobile processors because they have no IHS. Tcase is used for desktop processors.
 
Back