Topic: One question to anyone familiar with 3-4 GPU / OC and Gpu-z (solved) (Read 1266 times)
could be a higher vdrop on the last cards
April 20, 2013, 03:20:13 PM
tried but the computer crashed :O
And no way to know to which GPU i'm connecting ... (6990 : 1 DVI and 5 Mini Display Port)
gpu 1 is [0-0] and gpu 2 is [0-1]
And no way to know to which GPU i'm connecting ... (6990 : 1 DVI and 5 Mini Display Port)
gpu 1 is [0-0] and gpu 2 is [0-1]
April 20, 2013, 02:57:33 PM
I'm curious; would it be possible for you to try connecting your monitor to the other GPU? It would be interesting if your activity percent changed cards as well, or not.
Also, curious. Which of those cards does your miner register as 0,1?
Also, curious. Which of those cards does your miner register as 0,1?
April 20, 2013, 02:40:35 PM
6990 here so no PCI-E port in question. It seems my DVI link is attached to the second GPU.
The activity is lower on this GPU because of the desktop rendering.
(do not have Power Intensity Monitoring with this card :/ unlike you. Running GPU-Z 7.0.0)
April 20, 2013, 11:16:51 AM
Hey! Ty for reply
I cannot confirm that its an inaccurate reading per se, however it IS an artifice associated with the miner evidently.
Ye, run a Kill-a-watt.
Interesting your suggesting of the 96% relating to the display. ;P This was one avenues I explored as well. If you;
Plug monitor into GPU on slot 2 instead of slot 1 (e.g. so no display until windows OS kicks in). Slot 1 STILL shows the anomaly!!? even though the display is being provided by card 2. Heh, this is why it took me a while to finally figure it all out lol.
i would not know for the current seen on the first card. Probably inaccurate. Check with a Kill-A-Watt. But the 96% activity max you see is because the first card probably has your monitor attached to hit and the gfx card need to render the desktop ?
A computer cannot be headless. (has to have a graphic chip ...and render something ...)
A computer cannot be headless. (has to have a graphic chip ...and render something ...)
I cannot confirm that its an inaccurate reading per se, however it IS an artifice associated with the miner evidently.
Ye, run a Kill-a-watt.
Interesting your suggesting of the 96% relating to the display. ;P This was one avenues I explored as well. If you;
Plug monitor into GPU on slot 2 instead of slot 1 (e.g. so no display until windows OS kicks in). Slot 1 STILL shows the anomaly!!? even though the display is being provided by card 2. Heh, this is why it took me a while to finally figure it all out lol.
April 20, 2013, 11:05:21 AM
i would not know for the current seen on the first card. Probably inaccurate. Check with a Kill-A-Watt. But the 96% activity max you see is because the first card probably has your monitor attached to it and the gfx card needs to render the desktop ?
A computer cannot be headless. (has to have a graphic chip ...and render something ...)
A computer cannot be headless. (has to have a graphic chip ...and render something ...)
April 20, 2013, 10:04:54 AM
Ur um ok. I thought for sure THIS time someone would respond to my question. Oh well, still; ONE day..... ONE day a question WILL be answered, I just know it!
Anyhoo, solved this one too I guess. This is being "Caused" by the Miner. I would expect that GPU-z would show this artifice on any PC. I've been able to "relocate" which card shows the low amps, with various methods. Essentially whichever card gets listed First in windows will show this anomaly. I've also verified with a benchmark program that my Slot 1 (E2) CAN produce 100A, so again another confirmation.
Good thing is, I found a new trick and got my power down with a tiny bit more hash. All is well that ends well I suppose.
Figured I post the update in case any other used the Temp software and could know it's the miner.
Anyhoo, solved this one too I guess. This is being "Caused" by the Miner. I would expect that GPU-z would show this artifice on any PC. I've been able to "relocate" which card shows the low amps, with various methods. Essentially whichever card gets listed First in windows will show this anomaly. I've also verified with a benchmark program that my Slot 1 (E2) CAN produce 100A, so again another confirmation.
Good thing is, I found a new trick and got my power down with a tiny bit more hash. All is well that ends well I suppose.
Figured I post the update in case any other used the Temp software and could know it's the miner.
April 19, 2013, 07:06:05 AM
Ok, I have the Gigabyte 990FXA-GD80 Motherboard (I know many have also). 3 of the Powercolor http://www.newegg.com/Product/Product.aspx?Item=N82E16814131478
I have only run 3 for a couple minutes yet; waiting on risers. However, I've tested all the cards for setup in all possible combinations for all 3 cards now. I assume it's similar to other Mobo's for GPU plug in. (slot 1 for one card, slot 1,3 for dual and slot 1,2,3 for triple)
Here's the question and I cannot find an answer anywhere. No matter what GPU I test or combo of three for slot placement, one thing is always true;
With Gpu-z; The VDDC current of slot 1 NEVER exceeds 20A (not correct) max mining. However, any other slot (2,3,4) WILL run 100+/- amps (correct). As well, Slot 1 will not go over 96% gpu usage even OC, PT+20, and -I 20 on miner!
Sooo, why does slot 1 have such low current no matter which card? E.G. - is this normal for all your setups or maybe a bad mobo on slot 1 pci-e?
If anyone can help here it would be appreciated greatly!
Anyone with multi GPU and the software Gpu-z can test this to see also, I guess
http://www.techpowerup.com/gpuz/
EDIT: added GPU-z to show what I mean better. (PSU is 950w and whole system is 600w with 3 card)
Screen Shot of any of the 3 GPU in slot one; 96% max and 20A max.
Screen Shot of any of the 3 GPU in slot 2,3; 99-100% Max and 100A+
I have only run 3 for a couple minutes yet; waiting on risers. However, I've tested all the cards for setup in all possible combinations for all 3 cards now. I assume it's similar to other Mobo's for GPU plug in. (slot 1 for one card, slot 1,3 for dual and slot 1,2,3 for triple)
Here's the question and I cannot find an answer anywhere. No matter what GPU I test or combo of three for slot placement, one thing is always true;
With Gpu-z; The VDDC current of slot 1 NEVER exceeds 20A (not correct) max mining. However, any other slot (2,3,4) WILL run 100+/- amps (correct). As well, Slot 1 will not go over 96% gpu usage even OC, PT+20, and -I 20 on miner!
Sooo, why does slot 1 have such low current no matter which card? E.G. - is this normal for all your setups or maybe a bad mobo on slot 1 pci-e?
If anyone can help here it would be appreciated greatly!
Anyone with multi GPU and the software Gpu-z can test this to see also, I guess
http://www.techpowerup.com/gpuz/
EDIT: added GPU-z to show what I mean better. (PSU is 950w and whole system is 600w with 3 card)
Screen Shot of any of the 3 GPU in slot one; 96% max and 20A max.
Screen Shot of any of the 3 GPU in slot 2,3; 99-100% Max and 100A+
Jump to: