***Official***Z68X/P67A-UD7 Club/Discussion/Info**

Owi, did you clear the CMOS with the power unplugged after flashing?
This is probably the most important "Clear CMOS right after flash" BIOS.

Sin's CMOS is messed up because he has level 1, standard and level 2 still there...cuz he didn't clear the CMOS.

And the old LLC2 should be either #7 or #8 now, at least on UD5 and on the real_7's UD7. The real_7's UD7 software and DMM readings match my UD5 software reading.
The old LLC1 should be #4 (

Your board should not be dropping the vcore at load, it should be raising it.

Try clear CMOS, with power supply unplugged, then set LLC8 1.50v set in BIOS.

Quote:

---

Originally Posted by stasio

Thanks Sin,I am on level 4 now and uping.....
So far best BIOS for me. :up:
Some people over hardwareluxx.de forum talking nonsence! :rofl::ROTF:

---

Well i am doing another go at it, and it seems certain numbers from that chart are different. The reason being is that those results are from a B2 board using a different baseline LLC. Don't get me wrong many trends are the same, but I still think level6 is the best. I am retesting right now, and I am testing at 1.23v, 1.35v, 1.45v, and 1.5v and then getting idle and load for every level, hardware and software. I will post them without the fancy analysis numbers. For this minute, what i posted is a good rough indication of how LLC will work. Level6 doesn't have the same tight regulation i saw back a few weeks ago when i put that together for my own reference. I am in the process of changing it.

Quote:

---

Originally Posted by Falkentyne

Sin, ok you really did it this time.
I don't understand the chart :(
You have standard, level 1 and level 2, with ten levels in each one...how? :shocked:

Your BIOS has standard, level 1 and level 2, and with 10 levels in each one?
I think you need to clear CMOS with power unplugged, then?

Mine only has "Disable", and "llc1-llc10".
llc4 seems to be the same as the old "level 1", so I used LLC5.

---

Sorry man, yea i had a BIOS with the LLC levels, standard through level2 and then also that same BIOS lad LLC levels disable through 10, so I had 30 different LLC combinations.

Quote:

---

Originally Posted by owikh84

Tested F3 on my voltage loving 2600K at 5.0GHz
LLC10, Bios: 1.55, idle 1.65v, load 1.57v BSOD 124
Lower Levels are worse, cant boot into win.
Which level is equal to F3H Level2?

---

You are killing your CPU.

Owi's CPU is high VID, so its probably running cooler than other chips because it doesn't respond to voltage so well.

I think Owi needs to clear CMOS with power button unplugged because he is getting vdroop at llc10. vdroop should be removed at llc6 and rise about 0.05v at LLC8 ...

yea i see vrise occurring as soon as level 5. i think these levels change in practice. This might sound very weird, but i am getting weird results. This LLC is like adapting to different voltage levels. I stated above i would be testing each at 1.23v, 1.35v, 1.45v, and 1.5v. I am getting in some cases vdroop at 1.23v and 1.35v and then vrise at 1.45v and 1.5v. In other cases its opposite, and in others 1.23v vdroops and the others rise. Its pretty odd, but I think it follows a trend, the PWM is constantly compensating. Odd but its nice to see with different voltage levels. I am going to change what I said and make this preliminary statement. LLC level5 for voltages 1.4v+ is the best. I honestly think they dropped the LCL effectiveness down from what i did. I am still testing with wprime and now for this round I am running IBT at the same time, if I can still move my mouse around then its not loaded enough. lol j/k

All I gotta say is you guys are getting crazzy with all these LLC options.. I dont know if I should be jealous or not... Any chance this could be added to X58a-ud5 bios or not?

Sin...why haven't you cleared the CMOS?

I bet everything would be working perfectly and you wouldnt have all those mixed up options if you did :(

i did, but its not really needed. I accidentally set level5 and thought it was level6 lol. Anyways the PWm will compensate, so it will change per voltage, and those results are the same across the board.

These LLC options are pre programmed into the GPIO, but even more are also programmed. Maybe X58 BIOS has something like this, or it might be something new to the ISL63XX family.

I retested, like 7 times, i am coming to the conclusion that its just ripple and that others people's results might vary from 5 to 7mv, as mine are.

Quote:

---

Originally Posted by stasio

But,I am sure they improve LLC.
Test and see.:shrug:

---

stasio with the bios comming from B3 i need to flash with @BIOS and ALT+F12to backup i can load this F3 new one ? if anything goes wrong i can do @BIOS to F8f ?


btw anyone knows another vcore measure point ? as my waterblock is covering that zone you sin posted...

thanks :up:
Sergio

I updated the Bios for my UD7-b3 to F3.
My computer is totally unstable when vcore offset is activated.
I have infinite reboot loop. Finally I have to clear cmos to get it work.
Everything seems to be ok when vcore is fixed.

lol one of you wanna try something for me, with software that is, see if default gives you higher voltage than level1 and level2 and maybe equals level3 LOL

Quote:

---

Originally Posted by andressergio

stasio with the bios comming from B3 i need to flash with @BIOS and ALT+F12to backup i can load this F3 new one ? if anything goes wrong i can do @BIOS to F8f ?


btw anyone knows another vcore measure point ? as my waterblock is covering that zone you sin posted...

thanks :up:
Sergio

---

Any leg of any choke that is facing the right side of the board or CPU socket, but the MLCC capacitor reads a different voltage sometimes, the MLCC caps are all on the back of the board under the CPu socket.

Quote:

---

Originally Posted by sin0822

Any leg of any choke that is facing the right side of the board or CPU socket, but the MLCC capacitor reads a different voltage sometimes, the MLCC caps are all on the back of the board under the CPu socket.

---

yes sin but all the line of chokes is semi covered by the apogee XT i have to put a cable attached to my DMM if i want to measure but that sux i dont wanna touch anything else see ?

thanks :up:
Sergio

hmm you have a soldering iron? is the PWM covered too?

Quote:

---

Originally Posted by andressergio

yes sin but all the line of chokes is semi covered by the apogee XT i have to put a cable attached to my DMM if i want to measure but that sux i dont wanna touch anything else see ?

thanks :up:
Sergio

---

Dont be scared Sergio board and power supply have protection , even a slight touch and ground out system immediately will shut down , get your probes out :stick:

here we go, now this is a bit different than the first one, but i think you can figure it out. If someone wants I can also do idle-load to see the difference, but take a look and think about it lol. Level10 is a lot, level1 and level2 are worse than LLC disabled. Notice how far off software is. BTW your own measurements will be off anywhere from 5-10mv depending on your system. You can remeasure and you might even find that its 5mv off from your other measurement, its how VRMs work they are also compensating and changing the voltage. LLC level 5 and Level 6 are pretty good.
http://www.xtremesystems.org/forums/...1&d=1302411185

lol im never afraid of but i know where i can measure or not i will try my best

thanks real and thanks sin

does P67A UD7 B2 F8 BIOS implements this 10 LLC states ?

it might try it out? I think it might.

Quote:

---

Originally Posted by jacek

I updated the Bios for my UD7-b3 to F3.
My computer is totally unstable when vcore offset is activated.
I have infinite reboot loop. Finally I have to clear cmos to get it work.
Everything seems to be ok when vcore is fixed.

---

You are not alone. I run 24/7 @4.7GHz, LLC Disabled, offset +0.09.
Tried to optimize my Vcore a little bit with all those LLC levels with no luck. Any LLC level with offset causes infinite reboot loop and boots only after CMOS clear. Never had this problem with the old LLC options.

so i am guessing its just when you try to use the offset option?

EDIT: I just tried and you guys are correct, you cannot use offset and LLC infinite boot loop.

Time for another bios fix I guess.:)

yea probably a beta. For now just try without offset i guess.

I'm OK for now. Just can't get some fun playing with all those new levels for better 0-100% load Vcore scaling.:)
Wondering how GB made that big of move with no beta first.:confused:
Fire with high Vcore and LLC10 anyone?:D

Quote:

---

Originally Posted by andressergio

stasio with the bios comming from B3 i need to flash with @BIOS and ALT+F12to backup i can load this F3 new one ? if anything goes wrong i can do @BIOS to F8f ?

thanks :up:
Sergio

---

Yes Sergio,
if you crossflash B3 BIOS must use @BIOS(first time only),then Backup and all new B3 update with QFlash.
Going back from B3 to B2 BIOS,also must use @BIOS.
Btw,tested in both ways and no problem.
Always flash in "Optimized Defaults" and after CLEAR_CMOS.

Quote:

---

Originally Posted by jacek

I updated the Bios for my UD7-b3 to F3.
My computer is totally unstable when vcore offset is activated.
I have infinite reboot loop. Finally I have to clear cmos to get it work.
Everything seems to be ok when vcore is fixed.

---

Yeah,
this BIOS is different then last beta's,all my saved BIOS profiles DID NOT work in this BIOS.
Firstly,
I think that BIOS is defective,but latter on discover that must start from scratch and do all setting manually again (with CLEAR_CMOS and "Optimized Defaults").
So far,still testing (as much as I can).no reboot,no coldboot and adjusting optimal settings for various OC.
Backup BIOS still on F3h.

I know i said it before and before . . . . . . BUT I LOVE THIS NEW BIOS :)
it hard to understand if you dont understand you voltages between load idle and bios and what llc option do , but if you do understand this bring you so much closer to having your best load and idle close as possible so far Im on a high voltage 1.480 for testing purposes and using llc5 giving me a 1.428 load from dmm tester , so far great passed 1 hour prime , the combination voltages or endless :up: