Discussion in 'Videocards - AMD Radeon Drivers Section' started by AMDMatt, Apr 13, 2015.
what PSU do you have 11.66 it should be on 12.XX
No it shouldn't.
As the forum's my public profile listed,it's SeaSonic G-SERIES 650W.
But while GPUz reports 11.66,AB reports it is 12.19v.
Hi,ok I see.
But it seems I have no options to adjust voltage settings in AB.I already unlocked voltage control,If I am missing something with AB can you tell me how to achieve it?
GPU voltage:1299mv is already the top value AB allow me to adjust.So I have no way to set it more than 1299mv(ex:1300mv).
Memory voltage: I do no see any option allow me to do that.
As this Pic. shows:
I only have "core voltage/Power Limit/Temp. Limit(gray out)/Core Clock/Memory Clock/Fan speed. in AB.
I know there are options on other vender's cards(ex:MSI's),but it seems this card can't have the options.
Also I would like to ask another question. Does the GPU value that AB and GPUz reads them in different way?
Like the Pic. shows,AB reports I have 12.19v for core voltage,but GPUz says it is 11.99v.
Also sometimes GPUz will reports strange minus value (ex:-72v or something like that,that I can't believe what I saw), is that normal?
I heard that GPUz's reading is not really reliable, can I ignore this strange value or I'd better RMA it again?
I tested 15.4beta full day but this time still no TDR for now,I will keep trying anyway.
Yeah you might have a pwm controller which isn't supported or can't report voltage.
The GPU-Z 12v reading is from the pci-e line and read from the card. If u had AIDA or some other software you'd see another 12v reading which is taken from the motherboard.
So if someone wanted to eliminate the PSU as the problem they would check the board reading against the one on the GPU.
Low mobo 12v would likely be a bad PSU, if the board is OK and card 12v is low then either the GPU isn't connected directly to the PSU or there's some other fault but it lies with the card.
1. is on the motherboard, 2 is on the GPU.
Software readings aren't 100% accurate but close enough to be a good indicator.
As mentioned above your 12v input is a bit low but still reasonably acceptable.
Powercolor + overclocked GPU + Elpida Chips + overclocked vram + some unknown? pwm controller + it's been RMA'd once already.....
Sounds like unhappy adventures in PC gaming to me...
If the fault is intermittent which is what you've described then it may be coincidence that it hasn't crashed with the new drivers, give it some time to reappear.
I have a feeling you'll be seeing blackscreens again pretty soon, but we can hope. If it happens again RMA it and ask for a refund as I suggested earlier.
The gpu voltage in AB is the requested power, not what the vrm's are actually putting out. So no AB isn't giving a true reading.
That does not look healthy. I'd send the screenshot to Powercolor.
This is what healthy looks like
Driver seems to cause artifacts in GTA V with MSAA activated.
Here's a post + screen
Scaling is a bit off for me in GTA V. 50% 30% doesn't seem right
Coooler master 1000W 12v line idel 12.16 load 12.09
bios Mode 7970 to 1200/1650
In fact I sent the screenshot to Powercolor too when I RMA the card.But when the card sent back to me,this reading is still there.
It happens frequently when High Loading(not happens when low loading) although when it happens my PC is still fine and keep running:
Well,since the card does not black-screen freeze crash so I wonder it is fixed and did not ask them(Powercolor) about this issue for more detail.
I heard that software reading is not 100% accurate so I wonder if the minus reading is normal. Now I know I should not ignore that.
The card is connected to PSU's 6pin power connector directly.
And this minus reading also exists before I bought this G-650W,at that time I have a about 4 years old 400W psu and same situation.So basically I already bought a new PSU.But black-screen freeze crash issue still exists till I RMA'd the card.
G-650W is a semi-modular psu but it has a 6pin/8pin power connector directly from pus not modular.
I considered Cooler master V700W,but finally I bought Seasonic G-650W since I do not need a >600W PSU.
(15.4beta)I noticed similar issue when I am playing Blade&Soul.I have FXAA enable.
I got some white artifacts and then disappear, but it happens too fast to capture a screenshot.It only happened once.
What do you mean it doesn't crash? Earlier you posted a screenshot earlier of a TDR. Are u not getting any errors at all now?
If the ripple in GPU-z is accurate I'd say the vrm's are fubared, or maybe the transistors. But I said before to check the 12v on the mobo as well.
TDRs happen when a GPU command has taken too long to complete or the hardware is hung
The system does not reboot, but in most cases the screen flickers as it is redrawn. However, some older Microsoft DirectX applications render a black screen at the end of recovery, and users must restart these applications. These GPU hangs are referred to as time-out detection and recovery errors (TDRs).
So basically whether you see a TDR error or a blackscreen doesn't really matter, they're both caused by a stuck GPU.
Black screens often occur if the GPU gets stuck halfway through a driver reset.
When I said "not 100% accurate", I meant within a few mv... 12.19v in software may be 12.01v on a meter. So -70v on the 12v is not what I'd consider normal. Prob the pwm controller as I said before.
Can you just clarify one thing because this becoming unduly complicated.. Do u now have any TDR errors or blackscreens? Yes or no?
Before RMA(with old PSU and new G-65W PSU) : Only once TDR when I build this PC(old PSU), then random blackscreen freeze crash(means PC no responsding,only reset) for almost past 2 years.It is total random.
After RMA : With 14.12omega PC is fine
After 15.3/15.4 : I got TDR once then till today PC is fine with 15.4beta.
All issues are random so hard to find out how to repeat it.
As for Minus reading in GPUz,All happens before or after RMA.
OK so reinstall 14.12 and see what happens. If it still crashes then I guess u have your answer.
All that fluctuation in GPU-z doesn't look good though, I have to say.... I'll see if I can find a pic of the PCB
Hi, help please me ... i have a notebook with ATI Radeon Hd 5650 and i have update a driver with 15.4 ....
I play with 2 game : DragonBall and Pro Evolution Soccer 2015, but i have problem with pro evolution because when i start game i listen only and i cant watch :/ ... how can fix it ?
This problem occurs only when I install the latest drivers with Omega (14.12, 15.3 or 15.4)
Thanks in advance.
I just built another PC(well,in fact it is my old PC platform(AMD X4 630) and some old components that are not using now) for testing,moved my card to this test-PC and the same result, GPU-z reports minus reading after tested for a while in high loading.
So basically if there is any issue there, I know it should be on the card now.
Thanks for the advice, I will just RMA it again or buy another one to say goodbye to this card if TDR or blackscreen feeeze happens again, hope not
You guys seem pretty intent that the "blips" in gpuz indicate the card is faulty, unhealthy and needs to be RMA'd.... And there's no possible way it could be something more akin to what would be considered a polling error?
It might be actually, but there is a spike in one of the voltages during the blips. My theory still is it is the vBios.
i just can speak for myself. But i had the 7850 Powercolor for a while. And i had the exact same behaviour in gpuz. I think it's a reading error. I even had voltage peaks up to 12v. I never had issues with the card itself i just replaced it because the fan started to make noises. You shouldn't pay to much attention to the readings.
I just had a look in another forum (german) where i did a post about the behaviour. And it might be interessting for you i also got -70.51V on v12 back then.
I've got BSOD's when my pc goes to sleep with this driver.
Too bad, GTA V works fine.
COD Ghosts crashes
COD Ghosts crashes to black screen after every 1-2 multiplayer rounds. The issue continues from Omega to 15.4. AMD, please notice that bug. There are many complains about it on Steam, but of course nobody writes reports on the official site.
It's a polling error, I get the same thing in GPUZ at times. No issues with my card.