Overclock.net banner

1 - 16 of 16 Posts

·
Banned
Joined
·
9,168 Posts
Discussion Starter #1
My MSI board is being repaired (bad bios flash) so I'm stuck with a micro ATX ASUS board for the moment and I've noticed some odd things.

First my bios vcore is around .5 to .8 higher than my actual vcore in windows.

Second my temps seem a bit rediculious considering the vcore I'm getting in windows.

Is it possible my mobo is heating up at the vcore level in bios, but my cpu is only getting the vcore level in windows and getting less clocks with more heat because of it?

I can do 3.9MHz @ 1.392 vcore in windows, however that requires 1.45 vcore in bios to hit. I haven't fine tuned this level since I'm just dinking until my MSI board comes back but I find it odd.

To hit 4GHz I need to bump my bios vcore to nearly 1.6v which gives me around 1.52 in windows. I was able to run a little over 4.1Ghz on my other MSI board @ 1.47 vcore with way, way less heat than I'm getting now.

My solid state things that control the voltage on my ASUS board are getting insanely hot, I had to place a small fan over them because they would cause my system to shut down before the CPU would hit it's thermal threshold.

I bought two new 120mm case fans and have modded them onto my Zalman, temps right now @ 1.392 (1.45 in bios) are peaking @ 54 during LinX doing 2048MB tests.

Both my HHD's are sitting at 27C, so ambient is probably around 75F?

I'm itching for a few HSF, but I'm thinking most of my problem has to do with the MOBO pumping way more juice than it should be and the CPU isn't getting all of it. Should I wait for the return of my MSI board before I pull the trigger on a thermalright silver arrow?
 

·
Banned
Joined
·
9,168 Posts
Discussion Starter #3
I have no idea, I sent it back on 9/15. I assume it won't take much to repair, it just either needs the bios flashed or a new bios chip.
 

·
Registered
Joined
·
2,138 Posts
Quote:


Originally Posted by BallaTheFeared
View Post

My MSI board is being repaired (bad bios flash) so I'm stuck with a micro ATX ASUS board for the moment and I've noticed some odd things.

First my bios vcore is around .5 to .8 higher than my actual vcore in windows.

vDroop. Its normal.

Quote:


Originally Posted by BallaTheFeared
View Post

Second my temps seem a bit rediculious considering the vcore I'm getting in windows.

Maybe a reseat?

Quote:


Originally Posted by BallaTheFeared
View Post

Is it possible my mobo is heating up at the vcore level in bios, but my cpu is only getting the vcore level in windows and getting less clocks with more heat because of it?

I'm too braindead to understand?

Quote:


Originally Posted by BallaTheFeared
View Post

I can do 3.9MHz @ 1.392 vcore in windows, however that requires 1.45 vcore in bios to hit. I haven't fine tuned this level since I'm just dinking until my MSI board comes back but I find it odd.

I see no reason to complain. It took me 1.59v to hit 3.9 with an M4A79T Deluxe.

Quote:


Originally Posted by BallaTheFeared
View Post

To hit 4GHz I need to bump my bios vcore to nearly 1.6v which gives me around 1.52 in windows. I was able to run a little over 4.1Ghz on my other MSI board @ 1.47 vcore with way, way less heat than I'm getting now.

See above.

Quote:


Originally Posted by BallaTheFeared
View Post

My solid state things that control the voltage on my ASUS board are getting insanely hot, I had to place a small fan over them because they would cause my system to shut down before the CPU would hit it's thermal threshold.

They're called mosfets. And how hot they get depends on a few things. Heatsink, air flow, thermal contact etc. MSI uses Dr. Mos which is a very power efficient way to control the mosfets and reduce heat.

Quote:


Originally Posted by BallaTheFeared
View Post

I bought two new 120mm case fans and have modded them onto my Zalman, temps right now @ 1.392 (1.45 in bios) are peaking @ 54 during LinX doing 2048MB tests.

54 is fairly good temps for a Phenom at 3.9

Quote:


Originally Posted by BallaTheFeared
View Post

Both my HHD's are sitting at 27C, so ambient is probably around 75F?



Quote:


Originally Posted by BallaTheFeared
View Post

I'm itching for a few HSF, but I'm thinking most of my problem has to do with the MOBO pumping way more juice than it should be and the CPU isn't getting all of it. Should I wait for the return of my MSI board before I pull the trigger on a thermalright silver arrow?

Why? You're temps are fine and your overclock is fine. And you have a fairly well batched CPU.
 

·
Banned
Joined
·
9,168 Posts
Discussion Starter #5
Quote:

Originally Posted by TheOcelot View Post
vDroop. Its normal.

Maybe a reseat?

I'm too braindead to understand?

I see no reason to complain. It took me 1.59v to hit 3.9 with an M4A79T Deluxe.

See above.

They're called mosfets. And how hot they get depends on a few things. Heatsink, air flow, thermal contact etc. MSI uses Dr. Mos which is a very power efficient way to control the mosfets and reduce heat.

54 is fairly good temps for a Phenom at 3.9



Why? You're temps are fine and your overclock is fine. And you have a fairly well batched CPU.
This is just too much vdroop, the MSI board had hardly any.

Already reseated, AS isn't free


.15 vcore for 100MHz? qq?? It's also requires nearly 1.6vcore in bios, that's just too much


The heat becomes insane when the vcore is that high, or higher for 4.1GHz (1.55vcore windows 1.6250 bios)

Not the mosfets, the solid state capacitors next to them. I dare not touch the mosfets though, burns are no fun


3.9 is fine, but 4.1/2 is finer
 

·
Banned
Joined
·
9,168 Posts
Discussion Starter #6
I tested something, not sure if it matters or if it means anything, however here are my results.

a) random vcores upon entering windows, slight range difference with the same vcore setting in bios, albeit never higher, always lower if any change.

1. 1.43 vcore = 1.39 in windows
2. 1.44 vcore = 1.39 in windows
3. 1.45 vcore = 1.39 in windows
4. 1.46 vcore = 1.41 in windows
5. 1.47 vcore = 1.41 in windows
6. 1.48 vcore = 1.41 in windows

I never went any higher than this. Each step up in vcore raised my temps using LinX by 1C on average. However stability never changed when going for 4GHz, LinX would always error around the 5th or 6th test.

My thoughts are, temps are going up, voltage to cpu isn't.

What do you guys think?
 

·
Registered
Joined
·
1,069 Posts
Use a tool like PhenomMsrTweaker to check your power settings.
Windows has automatic power changing that can and will alter your voltages if you let it, Newer OS newer hardware newer hidden power management things etc.

I can set the CPU voltage to 1.2 volts in the bio's but without a tool to edit the windows power states windows will automatically raise this to w/e the cpu/chipset drivers have told it are proper or are default for the AMD cpu family group.

Most of these power changing "features" are inherently for the better. But can provide confusing results while OC'ing past Stock/Recommended settings.
 

·
Banned
Joined
·
9,168 Posts
Discussion Starter #8
I am currently using PhenomMsrTweaker, except it doesn't change it's voltage biased on bios, it's always at 1.3. If my vcore is 1.36 in bios, when I open up the tweaker it's 1.3. When it's 1.6 in bios I open up the tweaker and it's 1.3
 

·
Registered
Joined
·
566 Posts
Either CnQ is turned on or Windows power management is not set to high performance. Setting vcore in bios only affects one state (the max speed state, p0) in general.
For example, I set mine to run 3.2 at 1.175v in bios, but if I have Windows power management set to power saving, it jumps between 0.8 (default 1.225v), 1.6 (default 1.275v), 2.4 (1.325v), and 3.2 (default 1.375v). So whenever my cpu isn't running at max speed, the vcore will be read to be higher than what I set in bios.

Also, PhenomMsrTweaker tends to do strange stuff on my desktop (setting one core clock to be 157mhz, one at 400, rest randomly from 800 to 3.2ghz
) whenever I restart.
Try k10stat. There's a guide floating around somewhere.
 

·
Banned
Joined
·
9,168 Posts
Discussion Starter #10
CnQ is disabled

There are no options in windows power management to change the cpu min/max power settings. It' currently set to high performance.

k10stat did the same thing.

When over/under voltage protection is on my vcore is pretty close to where I set it in bios when I enter windows, however as soon as I load the cpu the voltage drops down to where it is when I turn off o/u voltage protection.

After LinX does a pack and the cpu goes to a more idle state the vcore jumps back up, soon as the load starts, it dives back down.
 

·
Banned
Joined
·
9,168 Posts
Discussion Starter #11
Just to give you an idea, right now it idles at 1.440 vcore.

Soon as I turn on LinX and it starts to work it drops to 1.392 vcore.

Soon as you finish a pack it jumps back up to 1.440 vcore.

/shrug
 

·
Banned
Joined
·
9,168 Posts
Discussion Starter #13
I got the answer I was looking for in the other thread.

I set my ACC to 0% and disabled over/under voltage protection.

Now my bios is matching my windows and the only thing left is to deal with the insane amount of heat I get at 1.4 vcore
 

·
Banned
Joined
·
9,168 Posts
Discussion Starter #14
Wow, that was scary


1.39 gave me about 54C before @ 4GHz.

1.42 showed 53C then 70C and I almost dropped a load in my pants


wth
 

·
Banned
Joined
·
9,168 Posts
Discussion Starter #16
Just an update on this, I'm at 1.312 vcore, 3998 MHz in windows, temps are starting to look better, 34/35C idle right now.

I'm confused


Edit: Wow, ok I'm not understanding this.

1.296vcore passed 1 test of LinX now, temps hit 61C 4GHz core clock x4

Going to drop my vcore a bit more, and a little less NB (1.23v atm) and see what happens.

Can anyone explain this?
 
1 - 16 of 16 Posts
Top