cancel
Showing results for 
Search instead for 
Did you mean: 

A few question about Ivy Bridge OC...

m4x2012
Level 8
Hi guys,
I want to know a little more about OC of my 3770k:



This was OK within yesterday, today all test that I conduce (IBT and LinX) will fail after a short time from start but with no BSOD, freeze or WHEA.

I've updated my M5G to lastest BIOS with "light" settings:

- Manual x 45 multiplier
- Manual CPU voltage 1,195
- Manual RAM voltage 1,50V and 1866Mhz profile (not XMP) 11-13-13-35
- HT and C3/C6 status Disabled
- Thermal protection off, LLC set to Ultra High with 140% Power.

Here is my system specs:

i7 3770k cooled by H100 Corsair
16Gb(2x8Gb)G.Skill TridentX 10-12-12-31 1,65V
Asus Z77 Maximus Gene V
Gainward GTX680 Phantom 2Gb
SSD Samsung 830 128Gb + HD WD Caviar Black 1Tb 64Mb
CoolerMaster Silent Pro 1kW
10,509 Views
18 REPLIES 18

HiVizMan
Level 40
I would look at your voltage Vcore after running memtest86+ on the ram. Use the USB stick version 4.2

Might need just a touch more voltage not much more I do not think but just a little.
To help us help you - please provide as much information about your system and the problem as possible.

HiVizMan wrote:
I would look at your voltage Vcore after running memtest86+ on the ram. Use the USB stick version 4.2

Might need just a touch more voltage not much more I do not think but just a little.


Thanks for your reply man.

Unfortunatly, memtest86+ says no errors on both sticks (tried each one alone).

I've adjusted @ 1,20Vcore and @1,55 DRAM voltage but issues persist.

My question is: Why the first time the IBT will pass without errors? 1,195 is a low voltage for 4,5Ghz but why the first time passed?

HiVizMan
Level 40
Ok so we know that your ram works individually. Please now try in dual channel mode at the frequency you are overclocking them too.

You can leave your CPU overclocked too that will give a more accurate result.

As to why it worked one day and not the next, well it is electronics and that is what happens some times. I have tested systems for a show on the Friday and they worked perfect, but would not even boot to BIOS on the Saturday. No reason at all, just electronics. Took them apart and put them back together and all worked as before. Go figure.
To help us help you - please provide as much information about your system and the problem as possible.

HiVizMan wrote:
Ok so we know that your ram works individually. Please now try in dual channel mode at the frequency you are overclocking them too.

You can leave your CPU overclocked too that will give a more accurate result.

As to why it worked one day and not the next, well it is electronics and that is what happens some times. I have tested systems for a show on the Friday and they worked perfect, but would not even boot to BIOS on the Saturday. No reason at all, just electronics. Took them apart and put them back together and all worked as before. Go figure.


Thanks for your help man.

I try asap your suggestion, but I can set my ram to the maximum profile (2400mhz with Auto voltage and timings) or it need to have a corrispondency to CPU Voltage/Clock?

HiVizMan
Level 40
With Ivy your memory and CPU frequency are independent of each other. I say this as you are not changing the BCLK to clock as your CPU has an unlocked multi. So no you can change your memory without having to do anything to your core frequency of the CPU.
To help us help you - please provide as much information about your system and the problem as possible.

X-ROG
Level 15
For 4.5GHz you shouldn't need to disable HTT or set 140%.

What about VCC voltages?

MarshallR@ASUS wrote:
For 4.5GHz you shouldn't need to disable HTT or set 140%.

What about VCC voltages?


VCC is set to Auto.

SoF
Level 7
somehow IB have the strange behaviour to do it's best oc on the first boot, then going whatever lets say 50 MHZ lower OR need a bit higher vcore. I have seen that on 5 GHZ as well as 6.8.

You can try to open the box of the pandora or just give 1.25V (or 100 MHZ less whatever) and let it run.
Maybe LLC was different during thoose runs, sometimes that's all it takes.

1.2V is very low for 4.5 anyway - had maybe 2/5 chips doing that ^^
now benching on: LN2 | 3770K | Maximus V Extreme | 2x GTX 580 Matrix Platinum | SSD | Cooler Master 1500W
Proud to be ASUS PowerUser!

SoF wrote:
somehow IB have the strange behaviour to do it's best oc on the first boot, then going whatever lets say 50 MHZ lower OR need a bit higher vcore. I have seen that on 5 GHZ as well as 6.8.

You can try to open the box of the pandora or just give 1.25V (or 100 MHZ less whatever) and let it run.
Maybe LLC was different during thoose runs, sometimes that's all it takes.

1.2V is very low for 4.5 anyway - had maybe 2/5 chips doing that ^^


According to my VID, stock voltage with default settings will be 1.26 (:eek:), so I'm working within this value.

However, I notice that increasing my RAM frequency will report errors both in IBT and LinX if I do a speed test in regular or high settings.
Repeating the same test w/o RAM increasing will be a +1.

Very strange thing, 200$ of RAM's certified with Gene V will do this issue.
I need to downclock manually to 2100 and set handmade timings and other data.