3770k stuck at 4.3GHz

[cbuchach]

I also have a 3770K and have been stuck around 4.3 (have it set to 43*100.5 to be exact). This is on an Asus Gene V with +.055 voltage offset(LLC off). 4.4 is borderline stable and 4.5 is a no go.

 

[Kenmitch]

I also have a 3770K and have been stuck around 4.3 (have it set to 43*100.5 to be exact). This is on an Asus Gene V with +.055 voltage offset(LLC off). 4.4 is borderline stable and 4.5 is a no go.

Temps are good?

Did you try enable cpu pll overvoltage to see if it helps or not with stability?

No go as in no boot or just unstable uner light loads, etc.?

 

[cmdrdredd]

What cooling are you using as well? With the Maximus V Gene I need my LLC set to high or it won't boot into windows easily for me. You might need a lot more than +.055 to get 4.4 or 4.5Ghz. I needed .13 for 4.6Ghz and 4.5 needs +.075 with LLC on high both times to avoid excessive vdroop. Not every CPU will do stock voltage 4.5Ghz clocks.

cbuchach I took some screenshots of my BIOS. Maybe it'll help you and the OP in this case figure something out. (please forgive the multiple shots.)

 

[Piano Man]

Indeed.. was seeing how far i could go on auto.

I am able to do 5ghz with 1.37v And it passed 6 hrs of prime 95. with temps around 90c On Air.

What?

You got 5.0GHz on a 3770K on air and it ran prime 95 for 6 hours? I've never heard of anyone going that high on air and being able to stress test it.


Can you post more info on your rig? I'm really interested to see what you got!

 

[cbuchach]

Temps are good?

Did you try enable cpu pll overvoltage to see if it helps or not with stability?

No go as in no boot or just unstable uner light loads, etc.?

Temps are OK. Will get just under 90C under maximum stress with Linpack. I have a COOLER MASTER GeminII S524 120mm heatsink. This is about the tallest HSF that will fit in my mATX case (I had an old TRUE in my old system).

I have not messed with cpu pll overwiltage yet, but will try it based on the thread comments here.

As far as stability, will certainly boot at 4.5 but Linpack will quickly crash.

 

[cbuchach]

Thanks A LOT for the BIOS settings, will help a great deal!

 

[cbuchach]

Quick question, I just checked my cpu pll voltage and noticed that with the auto setting, it is set at 1.80V and the sensor reads as such--where as CMDDRDREDD set his to 1.7125. What is the typical default?

 

[cmdrdredd]

Standard is 1.8v. Lowering it a bit from 1.8v can give a little better temps i've read. So I dropped it a bit.

 

[Kenmitch]

Quick question, I just checked my cpu pll voltage and noticed that with the auto setting, it is set at 1.80V and the sensor reads as such--where as CMDDRDREDD set his to 1.7125. What is the typical default?

The stock setting for PLL is 1.80v's raising it or lowering it may or may not help stability. Would be a try it and see tho....BSOD 0x124 sometimes is from PLL voltage if you set it too low for your chip or even too high.

Changing PLL voltage doesn't effect the need for Internal PLL overvoltage

Best description I could find about Internal PLL overvoltage so far. Kinda vague but gives a person an idea of what it does at least.

CPU PLL Voltage Override (Overvoltage): What the Heck does it do?
So I asked that question to an Intel Overclocking Engineer his explanation was roughly: We went through the BIOS settings trying to find setting that if changed could help overclock our CPUs further. We came across this setting. Think of the CPU PLL voltage as a voltage that is provided to the CPU, but then “clipped” down to an approximate voltage. No matter what that input is whether 1.3v or 1.9v it is clipped (hypothetically let’s say 800mv after clipping (he didn’t say how much)) that way other devices can use the PLL voltage and clip to what they need. The CPU PLL Overvoltage allows for less clipping of that voltage. It can also reduce the lifespan of the CPU, but nothing noticeable.

 

[Abwx]

The stock setting for PLL is 1.80v's raising it or lowering it may or may not help stability. Would be a try it and see tho....BSOD 0x124 sometimes is from PLL voltage if you set it too low for your chip or even too high.

Changing PLL voltage doesn't effect the need for Internal PLL overvoltage

Best description I could find about Internal PLL overvoltage so far. Kinda vague but gives a person an idea of what it does at least.

The PLL comprise a (low) frequency reference provided by a quartz
oscillator whose output frequency is sent to a phase comparator
who compare it to the frequency provided by a VCO (voltage controlled
oscillator) after it has been divided by a programable divider wich is
built with mos technology and whose ability to divide higher frequencies
is proportionnal to the power it consume , hence the PLL voltage setting.

 

[cmdrdredd]

I've been doing a bit of messing about in my BIOS, attempting to get more than 4.6Ghz stable at reasonable vcore. Might not be able to do it, but I'm gonna mess with LLC and PLL voltage. Apparently with LLC off you may be able to squeeze more Mhz into a lower vcore. I tested this myself and I'm stable at 1.26v 4.5Ghz with LLC off (need +.15v offset to account for droop). With LLC set to high I needed 1.275v. Not a big difference but it gives a little room.

Going to try messing with increased PLL voltage to see if I can get anywhere that way at less than 1.35v.

edit: well, I couldn't eliminate WHEA errors without a lot of voltage. So back to 4.5

 

[Abwx]

Anyway , be carefull when increasing supply voltages.

There s a threshold at wich current leaks start to increase dramaticaly
in some parts of the CPU leading to failure even if power comsumption
yield an apparently still safe temperature.

 

[borisvodofsky]

I've been doing a bit of messing about in my BIOS, attempting to get more than 4.6Ghz stable at reasonable vcore. Might not be able to do it, but I'm gonna mess with LLC and PLL voltage. Apparently with LLC off you may be able to squeeze more Mhz into a lower vcore. I tested this myself and I'm stable at 1.26v 4.5Ghz with LLC off (need +.15v offset to account for droop). With LLC set to high I needed 1.275v. Not a big difference but it gives a little room.

Going to try messing with increased PLL voltage to see if I can get anywhere that way at less than 1.35v.

edit: well, I couldn't eliminate WHEA errors without a lot of voltage. So back to 4.5

Um..... Lowering the pll voltage usually improves stability. This is the trend in the asus board forums at overclock.

I am running 1.65v

I tried 1.9, Has no effect.

 

[cmdrdredd]

Um..... Lowering the pll voltage usually improves stability. This is the trend in the asus board forums at overclock.

I am running 1.65v

I tried 1.9, Has no effect.

No...only lowers temps. Placebo effect. Raja told me this on xtremesystems

 

[borisvodofsky]

No...only lowers temps. Placebo effect. Raja told me this on xtremesystems

Who's raja?

And does lowering voltage lower temps, or does Higher pll voltage lower temps.D:

 

[cmdrdredd]

Who's raja?

And does lowering voltage lower temps, or does Higher pll voltage lower temps.D:

Asus tech guy from Taiwan.

He says PLL voltage can decrease temps by a small amount when you lower it and mentioned some people think it's more stable. He said it's only placebo effect and only increasing voltage can gain stability, not lowering.

Now remember he is from Taiwan so english isn't his first language. It is possible he was thinking of vcore in terms of stability but I am certain he mentioned possible slightly lower cpu temps if you try lowering PLL voltage.

Anyway if your CPU will boot at 1.38v and 4.8Ghz but throw WHEA errors, I know for a fact that 1.5v and 2.1v PLL voltage doesn't do a thing. Vcore will.

 

[borisvodofsky]

Asus tech guy from Taiwan.

He says PLL voltage can decrease temps by a small amount when you lower it and mentioned some people think it's more stable. He said it's only placebo effect and only increasing voltage can gain stability, not lowering.

Now remember he is from Taiwan so english isn't his first language. It is possible he was thinking of vcore in terms of stability but I am certain he mentioned possible slightly lower cpu temps if you try lowering PLL voltage.

Anyway if your CPU will boot at 1.38v and 4.8Ghz but throw WHEA errors, I know for a fact that 1.5v and 2.1v PLL voltage doesn't do a thing. Vcore will.

About those Whea errors. Which ones are really bad again?

 

[cmdrdredd]

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Internal parity error


I always get Processor ID : 2

 

[borisvodofsky]

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Internal parity error


I always get Processor ID : 2

I've never seen that, does it show up in a special corner of event viewer?

 

[cmdrdredd]

I created a custom view and put WHEA errors in there.

It's in the event viewer if you click on Event Viewer (Local) on the left pane and then expand warning in the middle pane. It's called WHEA-Logger

 

[cbuchach]

I've been doing a bit of messing about in my BIOS, attempting to get more than 4.6Ghz stable at reasonable vcore. Might not be able to do it, but I'm gonna mess with LLC and PLL voltage. Apparently with LLC off you may be able to squeeze more Mhz into a lower vcore. I tested this myself and I'm stable at 1.26v 4.5Ghz with LLC off (need +.15v offset to account for droop). With LLC set to high I needed 1.275v. Not a big difference but it gives a little room.

Going to try messing with increased PLL voltage to see if I can get anywhere that way at less than 1.35v.

edit: well, I couldn't eliminate WHEA errors without a lot of voltage. So back to 4.5

I had noticed similar things with LLC. My question regarding LLC is that the voltage will ramp up as the CPU ups its multiplier (from 1600 MHz to 4300) when it is loaded anyways--so LLC seems somewhat redundant. And, I was able to get a better OC with it "off" as my temps were too high with LLC "on" with similar stability.