M5A99X Evo R2.0 Issue radical help.

Hi all, its been a long time since the last time I unlocked a 4º core on my old Athlon II x3 440 and oced it 3,0 @3,6 stable.
A couple of days ago I upgraded my mobo from gigabyte 770T usb3 and the athon II for a M5A99X EVO R2.0 and a FX-8350.
So I wanted to oc it a bit since I had good memories from my old computer.

Here's the issue:

Stock vcore is set to 1.365 / Being in bios it says use "+" and "-" to adjust Voltage. So I press "+" and yes of course, the voltage raises as it should.
BUT if I press "-" the voltage won't go lower than stock 1.365 which is annoying, and I would say weird as f***. I'm sad about it because I have spent like 5 hours and haven't find how to make it work.
I really want to be able to lower vcore for optimal overclock capabilities.

Some info:
I have disabled C&Q EI power save stuff, C6 state, etc. Everything related to power saving.
I have LLC enabled to "HIGH / 50%" for an acceptable vdroop.
So far I have it stable at @4,4 at stock vcore "since I can set lower value" 1.365 idle to 1.330 full load. The issue is that i can run primestable at 1.286 so I'm wasting voltage and some Cº temp.

PD: I'm able to set lower voltage in AOD or Asus suit II but after system restart it goes agin to stock 1.365....
Any sort of help would be appreciated!!!!
Thank you in advance, and sorry about my bad english.

Because that board likely has what is called Cpu Voltage Offset mode you likely cannot just lower the voltage below stock volts. But put the board in Offset Cpu voltage mode and - / minus offset mode and you likely can then lower the voltage to the cpu.

SEE setting offset cpu voltage in your user's manual.

Did you run the Prime95 stress test in "blend" mode? If not, you should. How long were those Prime95 stress tests run for? Is you low voltatge at 4.4 ghz configuration stable enough to pass a two hour long Prime95 blend test run? Something is not quite right here. Usually, people have to increase the vcore from stock in order to overclock the amount you have.

What are you cooling that CPU with? That motherboard only has 6+2 powerphase and most folks find it won't give a stable overclock with the 8 core FX CPUs when they are significantly overclocked.

Is Turbo Boost mode still active in bios?

^^+1 was thinking the same thing about power phase

In the long range aspect of things he might have an under-manned motherboard to supply power to the FX-8350, if he tries to go much beyond the 4.4Ghz mark as did "johan45" and "mandrake" with their 990FX EVO and PRO boards. But that is not yet his condition. He wants to lower the Vcore below default and it might be P95 Blend stable there for 2 hours and it may not. For now though he just wants less voltage to the cpu. Then all heck might break loose. Hehehe.
RGone...ster.

True but the writing is on the wall.....

trents said:

Did you run the Prime95 stress test in "blend" mode? If not, you should. How long were those Prime95 stress tests run for? Is you low voltatge at 4.4 ghz configuration stable enough to pass a two hour long Prime95 blend test run? Something is not quite right here. Usually, people have to increase the vcore from stock in order to overclock the amount you have.

What are you cooling that CPU with? That motherboard only has 6+2 powerphase and most folks find it won't give a stable overclock with the 8 core FX CPUs when they are significantly overclocked.

Is Turbo Boost mode still active in bios?

Click to expand...

Answering all you guys, first thank you for your fast response!

Sorry, it wasn't 1.286 but 1.296. I ran prime95 for exactly 1 hour. It was set a bit higher than 1.296 but with vdroop it was running prime at 1.296 stable.
I'm using the same air cooling my phenom b40 had, a am3 cooler master hyper n620 + artic silver 5 thermal compund.
One side of case is open and its placed on the ground.
Max temp in prime95 = 51ºC
I have disabled turbo core.
I just want to know if my motherboard is defective about the voltage issue, so I can try to go for warranty.
I tried and I set Off set mode instead of Manual mode for its voltage control, and there you can click a symbol "+" when you click it you can turn it into "-" or vice versa, but nothing happen it dosen't go lower when I click the symbol "-" funny that even in offset if I use keyboard "+" it raises the voltage....
Thankl you all for your replies!

You need to see and do what page #86 says to do to set CPU Voltage in the motherboard manual > E8044_M5A99X_EVO_R2.pdf.

By the way in offset mode when you click on the minus sign; you do NOT click on it again, but instead you key in CPU voltages and they are reduced by the amount you are keying in. They do not appear in that voltage position as negative numbers but rather to the left side of the voltage you are keying in. Often must save and exit to get the actual reduced voltage to show in the bios.

I'll check the manual PDF and I'll let you know if I was able to lwer vcore.

I'm just bussy now so I'll update info later.
I'm realised stock vcore for fx-8350 = 1.28x? Is this true? If it is, is my vcore 1.368 safe for a 24/7? will it degrade the chip?

Ther aren't any 8350's I know of yet that the stock V_Core for 4.0 is below 1.3 something depending on the chip. My stock volts are 1.375v for 20x multi or 4.0. If you want to know your stock volts make sure the CPU is at default in the bios, start windows and open CPU-z. Click the about tab and save text file. Then go to where your CPU-z .exe file is located and there should be a text file containing that info. The line with the 20x is your default voltage at stock.

P-State FID 0x1A - VID 0x0A - IDD 13 (21.00x - 1.425 V)
P-State FID 0x19 - VID 0x0B - IDD 13 (20.50x - 1.412 V)
P-State FID 0x18 - VID 0x0E - IDD 12 (20.00x - 1.375 V)
P-State FID 0x12 - VID 0x16 - IDD 10 (17.00x - 1.275 V)
P-State FID 0xC - VID 0x1F - IDD 8 (14.00x - 1.162 V)
P-State FID 0x5 - VID 0x29 - IDD 5 (10.50x - 1.037 V)
P-State FID 0x10C - VID 0x33 - IDD 4 (7.00x - 0.912 V)

Here's a link to a collection we have going comparing P-States.

Johan45 said:

Ther aren't any 8350's I know of yet that the stock V_Core for 4.0 is below 1.3 something depending on the chip. My stock volts are 1.375v for 20x multi or 4.0. If you want to know your stock volts make sure the CPU is at default in the bios, start windows and open CPU-z. Click the about tab and save text file. Then go to where your CPU-z .exe file is located and there should be a text file containing that info. The line with the 20x is your default voltage at stock.

P-State FID 0x1A - VID 0x0A - IDD 13 (21.00x - 1.425 V)
P-State FID 0x19 - VID 0x0B - IDD 13 (20.50x - 1.412 V)
P-State FID 0x18 - VID 0x0E - IDD 12 (20.00x - 1.375 V)
P-State FID 0x12 - VID 0x16 - IDD 10 (17.00x - 1.275 V)
P-State FID 0xC - VID 0x1F - IDD 8 (14.00x - 1.162 V)
P-State FID 0x5 - VID 0x29 - IDD 5 (10.50x - 1.037 V)
P-State FID 0x10C - VID 0x33 - IDD 4 (7.00x - 0.912 V)

Here's a link to a collection we have going comparing P-States.

Click to expand...

Thank you very much, I just got confused because I saw a picture somewhere in a OC guide of 8350 and its stock vcore was 128x but whatever. I haven't touched Vcore since I can't set a lower value in bios so far, and I haven't increased it, so makes sense that in my case 1.368.
It's stable at this voltage @4,4 so I'm pretty sure If I increase clock just a bit like 4,5-, 4,6 it will be unestable at this stock vcore.
My worries:

As I have read from many oc guides I have disabled all power saving functions for oc advatages,so the clocks are static at 4,4 with a constant vcore of 1.368 idle,
AND /1.344~1.356 under full load "LLC set to high".
Will the cpu die younger than expected since Q&Q is disabled and all the 8 cores are always running at 4,4? or it doesn't even matter lol.
Same about vcore, will it hurt in the future a constant vcore than the 1.0xx idle vcore with power saving features? I know these may be very dumb questions lol I just don't plan on buying a new cpu any soon...

Last big question, if there is nothing to worry about actual settings:
http://valid.canardpc.com/036hx4

How far can I go increasing vcore to achieve a higher overclock? I mean without damaging the cpu or reducing its lifetime.
Thank you one more time.

Theoretically it will shorten the lifespan of the CPU but probably not enough to fry it before you have long since moved on to something else newer. The most important thing in that regard is not the actual voltage but keeping it cool. You seem to be doing a lot of the right things but you will need to increase the frequencies and voltages in small increments and test with Prime95 while monitoring temps and voltages with HWMonitor (non pro version) after each change.

You will get better results if you overclock with both the core speed multiplier and the FSB frequency together rather than just the multiplier as you are now doing.

trents said:

Theoretically it will shorten the lifespan of the CPU but probably not enough to fry it before you have long since moved on to something else newer. The most important thing in that regard is not the actual voltage but keeping it cool. You seem to be doing a lot of the right things but you will need to increase the frequencies and voltages in small increments and test with Prime95 while monitoring temps and voltages with HWMonitor (non pro version) after each change.

You will get better results if you overclock with both the core speed multiplier and the FSB frequency together rather than just the multiplier as you are now doing.

Click to expand...

So, just to make it clear :E.

Running clocks always at 4,4 and with a constant stock vcore. Theoretically it will shorten the lifespan of the CPU.
Whats the main source of this? the constant voltage of 1.368 or the clocks not going idle mode and always running at 4,4?
Thank you, I'm asking so maybe I can try enabling some features and checking if OC stills stable.
I don't want my chip to die young :E

Once you have your chip stable, if you know your stock voltage you can use the offset option in the bios which will allow your chip to run normally. Enable all the green stuff and balanced mode in windows and the power will cycle up and down as needed by your machine. !.368v really isn't much to worry about by the way.
If you're talking theoretically as soon as you OC your running out of spec and the consequences are yours and yours alone. As long as you're within reason your fine. If you wanted to run with 1.5v to the CPu constantly that is where you'll see some wear and tear on the CPU. Just my opinion.
Your voltage like Trents said is relative to your cooling. If you can keep the chip cool you'llbe fine. If you were running it even at 1.4v and the cpu was constantly in the 60-70c range that's an issue. See what we're saying

I believe it should run as you said not over 60ºC.
Well in this case it won't go higher than 51-52 ºC after 1 hour prme95 blend mode. Thats the core temp, but what about CPU temp? I saw it going over 59 to a max of 62 so far... is that dangerous?
PD: I'll set bios these features so vcore will be lower while idle, cheers!.

No sounds about right.

I did what you said, enabled balanced energy mode in control panel "Windows".
Enabled back all power saving features "C1E, Q&Q, APM master mode, core 6 state, etc".
And now as you said, the vcore while idle is 0.924, while full load 1.355~1.368.
I'm worried about if it is still stable... so I'll rune prime95 blend mode and report later here, thank you.
PD: Just asking, do I lose performance with this? I mean the clocks going to 1.4 ghz idle won't affet aero performance or opening web browser time, etc?. and while running heavy games like far cry 3 will it run always at 4,4? that would be good.
Thank you in advance.

Any task, even a background task, that begins to load the processor will cause Cool N Quiet to release and the voltage and speed will increase appropriately.

DEFEAT.
Even disabling all power features and only enable C&Q, its not stable... SO I switched back to disabled all power saving features and "Maximus performance" In windows control panel and its stable again...
So I'll go this way, I hope AMD uses good materials on the fx proccessors and hopefully it won't die sooner than at leat 3 years, that was whay my athlon II x3 440 @ phenom b40 @3.6 lived with no problems at all, evne with higher voltage than stock.... So I believe constant stock vcore shouldn't harm hard, thas my opinion I hope im right lol I'm not a rich guy but I really feel the performance gain with this little oc and power saving features disabled, everything runs just faster.
Thank you all for your time and patient with me and my bad english
PD: From your experience guys is best to let HT 2600 and NB 2200 "Which is default in my mobo or fx series idk" Or its better to let them at the same clock?

1. There is nO way to know if overclocking will shorten the life of YOUR cpu or really by how much.

2. It has been estimated by those much smarter than myself that cpus should last about 13 years full time use. Overclocking was thought to shorten that life to about 11 years. If you plan to use the cpu over 11 years, do not overclock.

3. As has been said the ability to keep the cpu cool is important as well as how much voltage is being applied to the cpu at what temperature. Cool cpus resist being damaged with added voltage over extended time periods.

4. One of the reasons we don't really try and teach setting offset cpu voltage and running with all the green stuff on is because it takes a little trial and error testing to get the voltages to drop and come back on with cpu speed increase. I have all the green stuff except C6 and Turbocore and my cpu idles at 1400Mhz and jumps to 4800Mhz when loaded and passes 2 hours of P95 Blend mode at 4.8Ghz. It can be done but it is not do A and B and C will be the result. It takes balancing the voltages to speeds.