what does these names mean

[paco2013]

Hello guys i have a msi gaming pro motherbord, and is see these voltage,s names, but what do the mean, and most of all what is there deafault voltage its not showing

 

[LigTasm]

Not to be rude, but we can't see what you're seeing. Either write the names down, or better yet take a screenshot of the BIOS (usually if you put a USB thumbdrive in and press F12 or something you can screenshot it) or if nothing else will work just take a picture of your monitor.

I'm guessing you're seeing things like Vcore, VCCIO, System Agent, etc?

 

[GMProtoss]

Not to be rude, but we can't see what you're seeing. Either write the names down, or better yet take a screenshot of the BIOS (usually if you put a USB thumbdrive in and press F12 or something you can screenshot it) or if nothing else will work just take a picture of your monitor.

I'm guessing you're seeing things like Vcore, VCCIO, System Agent, etc?

Exactly, what's your purpose? What are you trying to achieve with your question?

 

[michalrz]

Please take pictures, (be sure to resize them first! to like 640x480 or something) post them and we'll try to take a guess at them.

Every board has subtle differences.

Like in one board some settings might be renamed. So pics, pics pics and SMALL yo

 

[paco2013]

ah sorry guys, forgot to write it down.

i mean what does these mean

cpu pll oc voltage
cpu sfr voltage
cpu st pll voltage

cant see the deafault voltage, and i know it has somthing to dow with overclocking

 

[michalrz]

PLL is related to ensuring proper timing between the clock signal that the CPU receives and the actions it needs to coordinate and perform until the next clock signal comes.

It's one of those settings I would adjust on a trial and error basis and actually check real performance (like with a synthetic benchmark - Linpack?).

There was a thread not long ago where an AMD overclocker experienced low performance despite high, stable clocks.
Turned out, the internal Hypertransport bus was behaving erratically and was losing a lot of time on error management - which was all happening on such a low-level that software can't or doesn't even bother monitoring it. Different device, but shows that clocks are not always the whole picture.

 

[paco2013]

PLL is related to ensuring proper timing between the clock signal that the CPU receives and the actions it needs to coordinate and perform until the next clock signal comes.

It's one of those settings I would adjust on a trial and error basis and actually check real performance (like with a synthetic benchmark - Linpack?).

There was a thread not long ago where an AMD overclocker experienced low performance despite high, stable clocks.
Turned out, the internal Hypertransport bus was behaving erratically and was losing a lot of time on error management - which was all happening on such a low-level that software can't or doesn't even bother monitoring it. Different device, but shows that clocks are not always the whole picture.

oke thanks, bumped the pll voltage up a notche and happy with the result, 4.6ghz with 1.310v set in bios, all stable in everething i do with the pc

 

[michalrz]

oke thanks, bumped the pll voltage up a notche and happy with the result, 4.6ghz with 1.310v set in bios, all stable in everething i do with the pc

I haven't had the chance to OC these; how high could you go without bumping PLL related Volts? I.e. what was your previous overclock?

 

[paco2013]

I haven't had the chance to OC these; how high could you go without bumping PLL related Volts? I.e. what was your previous overclock?

did not test much, just put the pll voltage to 1.22v not more, i never had a problem with overclock, mine goes easy to 4.7ghz never had it higher, only tried booting at 4.9ghz with 1.4v and run superpi.
so on this bord it did not bring me much more oc, only a bit less voltage

 

[michalrz]

did not test much, just put the pll voltage to 1.22v not more, i never had a problem with overclock, mine goes easy to 4.7ghz never had it higher, only tried booting at 4.9ghz with 1.4v and run superpi.
so on this bord it did not bring me much more oc, only a bit less voltage

That's awesome if you could get away with less 'other' voltages after bumping PLL. (unless I misunderstood you)

One more question - did the lower voltage correspond to a drop in load temperatures?

I'm asking because I remember reading some voltages are actually offset from their reported values - so that PLL bump might have actually raised your vcore also (but you'd have to physically measure it with a meter). I'm not familiar with newer chipsets though so I might be repeating some old wive's tale.

 

[paco2013]

That's awesome if you could get away with less 'other' voltages after bumping PLL. (unless I misunderstood you)

One more question - did the lower voltage correspond to a drop in load temperatures?

I'm asking because I remember reading some voltages are actually offset from their reported values - so that PLL bump might have actually raised your vcore also (but you'd have to physically measure it with a meter). I'm not familiar with newer chipsets though so I might be repeating some old wive's tale.

did not measure with meter, just what it shows in cpu-z, and hwinfo.
Temps defference is not much so i set it all back to normal, overclock to 4.6ghz/1.30v set in bios, during load that gives me 1.312v and after a bit testing i,m stable for my 24/7 overclock.
temps durring gaming ar 53 degrees for highest core.

i can test for hourse temp will not change

 

[michalrz]

Cool beans.
Guess I won't ignore PLL anymore lol

 

[paco2013]

Cool beans.
Guess I won't ignore PLL anymore lol

as you can see for 4.7ghz i have to set the vcore 0.30v higher 2 degrees in temp more, same benchmark same ambient temp of 20 degrees celsius

 

[michalrz]

Seems you're loading only one CPU core (I'm seeing ~25% utilization in Realtemp in both screens), is this intended?

 

[paco2013]

Seems you're loading only one CPU core (I'm seeing ~25% utilization in Realtemp in both screens), is this intended?

dont know, i think its a false reading, because if you look at the themps you see the are all high

look at this picture at 4.5ghz
you see also realtemp at 25%

 

[paco2013]

sorry taken after load so now its 1.3% but its a false reading of realtemp, dont now why

 

[michalrz]

sorry taken after load so now its 1.3% but its a false reading of realtemp, dont now why

I do hope it's just a false reading because you got a nice OC going. And the temps are beautiful.

However, a few centuries ago I was overclocking my Q9550 and realtemp was the first piece of software that suggested I was suffering from Voltage Regulator overload and some sort of throttling. It was stuck at 76%, even though task manager showed all four cores loaded to 100%.
I then ran Unclewebb's Throttlestop and - upon launching - the realtemp reading shot to 100%, my temps skyrocketed and after a few months of usage my Voltage Regulator (circuitry on the left of the CPU socket) made the motherboard warped and brown. Obviously the motherboard was doing its job and was preventing early death and it DID knew better...

Besides, 125 GFlops is quite on the low side for such a modern high-clocked CPU.

I see you're using the glorious hwinfo - good call.
I friggin' love that thing, it seems to know everything.

Could you please post the same screen but at full load? I'm genuinely interested - not trying to disappoint you or anything

 

[paco2013]

I do hope it's just a false reading because you got a nice OC going. And the temps are beautiful.

However, a few centuries ago I was overclocking my Q9550 and realtemp was the first piece of software that suggested I was suffering from Voltage Regulator overload and some sort of throttling. It was stuck at 76%, even though task manager showed all four cores loaded to 100%.
I then ran Unclewebb's Throttlestop and - upon launching - the realtemp reading shot to 100%, my temps skyrocketed and after a few months of usage my Voltage Regulator (circuitry on the left of the CPU socket) made the motherboard warped and brown. Obviously the motherboard was doing its job and was preventing early death and it DID knew better...

Besides, 125 GFlops is quite on the low side for such a modern high-clocked CPU.

I see you're using the glorious hwinfo - good call.
I friggin' love that thing, it seems to know everything.

Could you please post the same screen but at full load? I'm genuinely interested - not trying to disappoint you or anything

love hwinfo, i think the problem with realtemp is that its old, and same with itb, 120gflops is normel for 4.5ghz

the picture is taken under load

 

[michalrz]

Yup, you're right - all is good nice...

 

[Dan_D]

On Skylake you can usually ignore CPU PLL voltage settings. It doesn't normally net you anything more than "auto" does. For older Intel platforms adjusting this value can increase stability while overclocking.