Overclock.net banner
1 - 9 of 9 Posts

· Still kinda lost
Joined
·
3,912 Posts
Discussion Starter · #1 ·
As the title states how is this even possible??? i only changes the Multi and the Vcore,Closed all the power saving options and turned off the turbo boost and the rest of the voltage options @ auto, So any ideas what just happened? and the max temp i had was 72 in dota 2 which heats the cpu more than crysis 3 and startcraft 2 ( a CPU needy game as you know)


Idle temps on the left and on the right are Temps when i open programs like Chrome (Spikes for secs only)
 

· Programmer
Joined
·
29,887 Posts
Too much voltage can cause instability.
 
  • Rep+
Reactions: Awsan

· Registered
Joined
·
3,463 Posts
While more voltage can cause instability, you aren't really pushing high voltages so I think something else is going on.

My guess is that you are not properly stability testing to your chip, so you are not really stable at [email protected] Then you begin pulling your hair out when 1.35v seems 'more' unstable.

Unstable is unstable. This is why you really need to test for 24 hours prime95, sum error and rounding checking, max memory used, priority 10/task manager above normal. It doesn't matter if an instability doesnt come until the 31st hour or in 5 minutes. That can sometimes be a sign that an overclock is 'more' stable but not always, it just means that what the particular stress program is testing in the first 5 minutes hits a part of the CPU that's different than what it hits at the 31st hour.
 

· Still kinda lost
Joined
·
3,912 Posts
Discussion Starter · #4 ·
Quote:
Originally Posted by Belial View Post

While more voltage can cause instability, you aren't really pushing high voltages so I think something else is going on.

My guess is that you are not properly stability testing to your chip, so you are not really stable at [email protected] Then you begin pulling your hair out when 1.35v seems 'more' unstable.

Unstable is unstable. This is why you really need to test for 24 hours prime95, sum error and rounding checking, max memory used, priority 10/task manager above normal. It doesn't matter if an instability doesnt come until the 31st hour or in 5 minutes. That can sometimes be a sign that an overclock is 'more' stable but not always, it just means that what the particular stress program is testing in the first 5 minutes hits a part of the CPU that's different than what it hits at the 31st hour.
12 hours prime and its still going, ok lets presume to worse if its not stable @ 1.3 or 1.35 what should i do ?
 

· Programmer
Joined
·
29,887 Posts
How long did you test at 1.35v and how did you test.
 
  • Rep+
Reactions: Awsan

· Registered
Joined
·
838 Posts
Quote:
Originally Posted by awsan View Post

12 hours prime and its still going, ok lets presume to worse if its not stable @ 1.3 or 1.35 what should i do ?
If it's not stable at 1.30, try again at the next lower notch. If it's STILL unstable, it means you're almost definitely outside the the chip's OC capabilities.

So then what I'd do is put it back up to 1.35 and start lowering clockspeed incrementally until you reach a 24 hours stable Prime95 run. THEN, when you're 100% sure of stability, start lowering the voltage again back to 1.30 and see if it's still stable. If not, you know that X GHz @ 1.35 is your max performance attainable. Yes it's tedious, and a total pain, but welcome to our world!

Some chips just don't overclock as well as others, simple as that I'm afraid.
 

· Registered
Joined
·
3,463 Posts
Starcraft 2 isn't really going to heat up a CPU because it only uses 2 cores and is really only going to be burning things up in 200 vs 200 battle, by the way.

And dota2 isn't going to be heating things up either. You should do small fft prime95 for temp testing, which will be at least 20-30*C hotter than a game like sc2 or dota. Small fft prime95 is just fft length 8, which is the 2nd test pass in prime95 blend (ie minutes 15-30) so you should see what your max temp got to in your monitoring program.

As others said, if you aren't stable at a certain voltage, you adjust voltage/frequency as necessary. I had a few ivies that needed 1.45vcore for only 4.6ghz. Most ivies are in the 'bad' lot and overclock similarly, only 4.5-4.6 around 1.35-1.4vcore.
 

· Still kinda lost
Joined
·
3,912 Posts
Discussion Starter · #8 ·
In that matter i already tested [email protected] 1.3 for more than 24 hours on prime95 and some Intel burn test and it held up like a champ so i guess i will go with that until i find a perfect guide on how to use other settings like PLL and other Voltage settings, Thanks every one for the help + REP
 

· Registered
Joined
·
3,463 Posts
I don't know what you are looking for, i wish i could tell you there's some irrelelvant setting you could magically change but you cant. The other voltages and settings really arent necessary to mess with. Like they dont do anything.

You change vcore and frequency (and use offset voltage once you figure out your max overclock so as to keep power saving options and extend life of cpu and stress it less). Just have LLC to 2nd highest setting, 'enable pll overvoltage'. After you got your max voltage and frequency down you can maybe reduce the pll, vtt, imc, gfx voltages to reduce heat but it's not a big deal.

http://www.overclock.net/t/1247413/ivy-bridge-overclocking-guide-with-ln2-guide-at-the-end

Good guide on overclocking. Its not hard to find a good overclocking guide.

So in your case,you now got a baseline and know for sure your hardware is not faulty and all is good. So work from there, try 4.7 at 1.3, if not, maybe a small bump up. [email protected] is not bad, you can probably get a higher overclock with a bit more voltage, depending on how temps are and what you are comfortable with.
 
1 - 9 of 9 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top