Ivy Bridge Overclocking V1 | 3770K/3570K | Z77, Torture Your CPU, Post It Here!
|
|
 Apr 13 2013, 11:32 AM
Return to original view | Post
#1
|
 
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night  
|
Just got a 3770k Costa Rica. To my surprise, temp when running Linx at 4.4ghz @ 1.24v actually shot up to 90c. My proc is cooled by h100 with 4 gentle typhoon push pull configuration with arctic mx4 tim. Looks like h100 performs worse than some of the high end air cooler.
|
|
|
|
|
|
Apr 13 2013, 12:17 PM
Return to original view | Post
#2
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(owikh84 @ Apr 13 2013, 11:58 AM) No matter how good is your cooler, even a quad TEC, you wont be able to tame the Ivy Bridge because there is still a culprit underneath the IHS, until the IHS is delidded, black adhesive glue is removed and TIM is reapplied. Then only you will see the improvement. So guys, dont waste your money for expensive cooling gear, if you dont plan to delid your chip. Thank you for your advice. But i have seen some posts in this thread people using air cooler like noctua D14 and CM air cooler actually have better temp than mine. Could it be H100 pump failure? |
|
|
Apr 14 2013, 11:53 AM
Return to original view | Post
#3
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
OK, i have found out the reason for my high temp. It was due to improper application of TIM
 I too kedekut and applied only half a drop of TIM instead of a drop of a small size pea  Before 
After reapplication of TIM: 
Room ambient temp of 31c BTW pls add me to the list. Still finding out the lowest Vcore for my chip. This post has been edited by hypermax: Apr 14 2013, 12:06 PM |
|
|
Apr 15 2013, 01:56 PM
Return to original view | Post
#4
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
Funny things that I observed. My system passed Linx 25 runs and aida64 stress test > 12hrs, but failed prime95 large fft within 15mins. So it's my overclocking stable? I am now trying for 4.5ghz.
|
|
|
Apr 15 2013, 03:51 PM
Return to original view | Post
#5
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
Means I got a dud chip then. Needs such high voltage for a mere 4.4ghz overclock.
BTw, When I first overclocked my chip to 4.4ghz, I set my voltage to auto. Then it turned up that the stupid mobo put 1.47v to the vcore. The pc managed to boot into windows then shut down by itself. Just wanna ask Could the initial high voltage damage the chip and made it requires more vcore when overclocking? Sorry for the noob question. Still new to the overclocking world. This post has been edited by hypermax: Apr 15 2013, 03:52 PM |
|
|
Apr 15 2013, 08:17 PM
Return to original view | Post
#6
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(owikh84 @ Apr 15 2013, 04:09 PM) I have been using 5.0GHz @ 1.48v for 24/7, haven't seen any degradation since the first day. Thanks for the info. Ivy Bridge is a lot more durable than Sandy Bridge in which the latter is a hell of degrading craps. In conclusion, IB will not degrade at all, until you put an extremely high voltage on it say 1.9-2.0v.  Btw, how reliable is aida64? Asus opines that it is THE ONLY software which understands ivy bridge, thus passing aida64 stress test > 12hrs indicates stability. This is quite different from what I have learned previously. I always thought stability is only achieved by passing prime blend test 24hrs. Btw, I bumped my voltage to 1.3 from 1.288. The GFlops in Linx is still the same is my proc something wrong?This post has been edited by hypermax: Apr 15 2013, 08:18 PM |
|
|
|
|
|
Apr 16 2013, 07:16 PM
Return to original view | Post
#7
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
Still trying to figure out how to overclock my ivy bridge. It is a beast difficult to be tamed. Unlike sandy where u can just pump vcore and get stable overclocking, ivy bridge hates excessive vcore. The vcore has to be "just enough" for a overclock to be stable
I got a lot of d124 error which was due to excessive vcore.Also, I found other voltages like pll, vccsa, vccio actually play a role in the stability of the system even in modest overclock. This post has been edited by hypermax: Apr 16 2013, 07:20 PM |
|
|
Apr 16 2013, 09:42 PM
Return to original view | Post
#8
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(owikh84 @ Apr 16 2013, 09:27 PM) You may run LinX/IBT/prime95 in diagnostic mode, so that it won't be affected by windows background services and can use maximum RAM amount as possible as there would be more free mem available. When I was stressing with prime, I can't pass half an hour no matter how much vcore I put in (upto 1.42) at 4.5ghz. I always get the BSOD 124. Then I came across an article on hardocp which states that increasing PLL to 1.9 helps. I did so and lowered my vcore down to 1.32v. Now my prime is entering its 3rd hour of testing. Every chip is unique I guess.I can't comment on your experience but IMO Ivy is a lot easier to overclock than Sandy, provided you have a good cooling gear to tame its temps. Sandy is suffered from the infamous BSOD 124 (insufficient vCore) and deteriorating so quickly within a few days. I have never heard of the Ivy hates excessive vCore. hmmm... vPLL doesn't help in overclocking at all. I used 1.35v to 1.8v also won't give any different in terms of overclockablity and stability. VCCIO aka vIO = VTT and VCCSA aka vSA = IMC will help RAM overclocking ONLY, not CPU OC. Btw, I am now using 16gb for ram populating all ram slots. I find that increasing VCCIO from 1.05 to 1.1 actually helps to stable the system. No sure though since I am still very new to IVY bridge overclocking. This post has been edited by hypermax: Apr 16 2013, 10:49 PM |
|
|
Apr 16 2013, 10:47 PM
Return to original view | Post
#9
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(jimmylim85 @ Apr 16 2013, 10:22 PM) I have a problem. I can't get pass 4.2ghz even i increase vcore and cpu ratio to 45x.. what's going on? Use jumper to clear CMOS. I have this issue with my P8Z77 V as well. |
|
|
Apr 16 2013, 11:08 PM
Return to original view | Post
#10
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(jimmylim85 @ Apr 16 2013, 11:00 PM) Sounds like i have to the hard way by opening the cover and get a jumper to short the CLEAR CMOS? That's what I did and it worked for me.Did it work for you on this method? |
|
|
Apr 17 2013, 10:51 AM
Return to original view | Post
#11
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(jimmylim85 @ Apr 17 2013, 09:40 AM) Thanks, Is very safe. In fact u got a very good chip with low vcore. For me, I have to run 1.328v at 4.5ghz to be prime stable.Btw is it safe and stable to run 4.5ghz on 1.150 vcore? |
|
|
Apr 17 2013, 11:19 AM
Return to original view | Post
#12
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(owikh84 @ Apr 17 2013, 11:07 AM) Bought these strong PSC kits recently. Wa, KVR. I am interested to know the result of KVR Corsair GTX4 2533 C9  KVR 1333 C9  Stay tuned.  Anyway, finally got my system prime stable: 
Previously can't even run prime for half an hour even at 1.42v vcore However, to be Aida64 stable needs much less voltage. And this is before i raise the PLL: 
I will try to run my system at Aida stable voltage from now, see whether aida is reliable or not. |
|
|
Apr 17 2013, 11:31 PM
Return to original view | Post
#13
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(owikh84 @ Apr 17 2013, 11:02 PM) As promised, I'd would like share some results from the cheapo KVR 1333 C9 sticks, that might put some of the more expensive kits to shame HAHA Wow, KVR is amazing.  KVR 1333 C9 2 X 2GB 1.5v  2400 7-11-8-28-1T @ 1.685v  2500 8-12-8-28-1T @ 1.83v  Btw, may i know is your kit cherry picked or randomly grabbed? |
|
|
|
|
|
Apr 19 2013, 12:15 AM
Return to original view | Post
#14
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(Dreamworks @ Apr 18 2013, 11:31 PM) Reduced my PLL voltage from BIOS default of 1.85V (shouldn't it be 1.80V? Try increasing to 1.9v. Might help in reducing vcore. ) to 1.50V just now, doesn't seem to affect stability nor reduce temps. Only problem I've seen so far is a few apps like RealTemp and Intel XTU detecting wrong CPU clocks and wrong BCLK owikh84 How come my RAMs running at 1866MHz, C9, 1T doesn't seem that much slower than your 2400MHz, C7? Except for latency  [attachmentid=3399027] |
|
|
Apr 25 2013, 01:49 PM
Return to original view | Post
#15
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
New overclock. Managed to overclock at 4.5ghz with 1.3V. Vcore still a little high. I believe with a better board my vcore should be lower.
This post has been edited by hypermax: Apr 25 2013, 02:06 PM |
|
|
Apr 28 2013, 01:48 AM
Return to original view | Post
#16
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(souq @ Apr 27 2013, 10:51 PM) Hi all You are using the older prime which doesnt have AVX instruction. Thus your proc is not fully stressed. Pls use the latest Prime at 1st page of this thread in order to stress your cpu to the fullest.my first try to OC my 3770K is it ok at this temperature and voltage?, i7 has been delided, and cooled without metal shim, direct contact between silicon to waterblock (seidon 120M), push pull fan configuration.  please help tq |
|
|
Apr 28 2013, 11:34 AM
Return to original view | Post
#17
|
|
Senior Member
2,524 posts Joined: Jan 2003 From: Beneath the starry night
|
QUOTE(souq @ Apr 28 2013, 07:39 AM) ok thanks @ketapang, @hypermax.. i'll re stress with latest Prime.. sorry, noob here Bro, no need to apologize. We are all here to share and learn |
|
Topic ClosedOptions
|
| Change to: |  0.0481sec
 0.45
 7 queries
 GZIP Disabled
Time is now: 7th December 2025 - 11:50 AM |
Quote