Printable View
few min ago i was testing my card and find out something weird
the voltage i set was original 1.55v (vmoded) in 975
but when i measure it today in 680i it is like 1.65v ....
and even i remove the mod, the voltage reads 1.48v idle and 1.5v loaded....
i haven't test back in 975...
680i mobo is asus rog...
That's odd. 8800 don't have 2D and 3D profiles with a different vcore right? If it does have them that would be the only logical explanation.
Waiting for the re-test in the 975x board.
I`m getting different voltage readings on my GTS compared to everyone else.
I get 1.34V without voltmods, checked with 2 DMMs, way higher than anyone else.
Board is an Asus P5B DH
How does it overclock? Maybe Nvidia bins the dies to different vcores/revisions like AMD and Intel do?Quote:
Originally Posted by K404
625/975, but I dont like my temps...sorry, not pushing it until I can sort out my cooling
I know it`ll bench higher
ok it is for sure
i am running in on 975 now and core voltage is 1.55v and memory voltage 2.25v
in 680i core voltage is 1.64v and memory voltage 2.35v....
Hmmm thank's for info :toast:
This is good for 680i user, but generally strange.
btw : I moded my 8800GTX ( Asus ) gpu@200 Ohm and :slapass: Vgpu 1.5V with mod default - clocks 660MHz ( ac cooling temp stress 60C ) and what ? No more MHz, more vgpu do nothing. I tried 1.7V and complete nothing more :( Water on gpu and 660MHz :confused: Bad Asus card :stick:
Regards
Martin
Do you use the same ground for both measurements?
Different rails can have different ground levels, could be one explaination.
The vgpu ground plane can also float a little.
Try to measure the voltages in reference to the PCI-E ground on each card, and also compare them to the motherboard ground (screw holes and such).
Another thing I have noticed is that some voltage regulators use the 3.3V rail as a referens voltage. (A P5W64 overvolts the vcore higher with higher +3.3V for example) There might be a droop on the 3.3V voltage to the PCI-E ports on one of the boards.
Maybe nvidia did that on purpose to secretly make the 680i overclock cards better?
where do you have slot power @ ?
it is the same card and i didn't change / remove / remod the card tho...Quote:
Originally Posted by crotale
and tested with two cards
@Dynasty
slot power? on 975 is auto and for 680i i don't know there is slot power adjust
i use vcore measure point on the back of 8800GTX
I was thinking the same thing..Quote:
Originally Posted by Demo
i did call a nv card sellerXDQuote:
Originally Posted by NickS
he told me it is for sure 680i will increase the voltage of vcore and even with tighter mem timing...... well i think most of us has both 680i and 975 /965 mobo. plug your 8800GTX or GTS on it and measure it.
also, stock 8800GTX on 975 is 1.3v and in 680i is 1.5v....
and memory voltgae is higher in 680i too
do you get higher max 8800GTX clock on 680i due to the higher volts ??Quote:
Originally Posted by guess2098
yesQuote:
Originally Posted by eva2000
since i thought my card had problem lol, so i removed one card's vmod, and i reach 700 with "stock" air with artifact.... (1.5v)
i reached 710 with vmod on 975 with 1.55v (200ohm vmod started at 1.55v)
wow... do you know if same happens with 8800GTSs ??
i don't have GTS anymore..... so................................Quote:
Originally Posted by eva2000
i am waiting to see your GTS hehehehe
DO IT man
1.5VCore is obviously safe on stock cooler then! Tho temps in an overclocked state might be quite high...
I dont understand how can different board affects VCore on G80...
Maybe we will Vmod the graphic cards trough the system BIOS in the future :)
I noticed one thing, strange, unexplainable
8800GTX on P5B Delux mobo, 8800GTX with Shamino's vgpu & vmem mod
2D voltage = 1.75v
3D voltage = 1.65v
I am 100% sure that i measure this right, also with 8800GTS same thing
the same voltages also on P5W64
Maybe the 8800 cards do have different VIDs, and we don't know it? Could be that the mod is targeting the wrong VID, don't know why the mod effects both VIDs though. Ohh wel, I have no knowledge of the power circuity at all, so I have no idea really.Quote:
Originally Posted by zbogorgon
Different VIDs is the only explanation why the cards have a higher vcore on 680i boards. Mmm, there might be a way to let the video driver think the motherboard has a 680i chipset while it actually has an Intel one. For the higher vcore of course.......
Didn't nVidia advertise something about overclocking all parts ~5-10% if you used an entirely NV platform? I remember hearing something like this a while ago. If that's the case then pushing slightly more voltage into the card in a 680i mobo makes sense.Quote:
Originally Posted by Demo
I think you may have hit the nail on the head with that one.Quote:
Originally Posted by Demo
Have you tried measuring v's with Linkboost enabled and then disabled, could Linkboost be causing this overvolting :confused:
Linkboost and vgpu go up ? Hmmm not. For me linkboost = raise PCI-E freq.Quote:
Originally Posted by lowdog
Perhaps the option in BIOS (Nvidia GPU eX) causes the increase in volts?
My suggestion would be to boot bone stock and measure in 2d/3d
Then increase individual bios vooltages one at a time measuring 2d/3d volts on boot.
And finally if the voltage options dont affect it, perhaps trying various options, i.e. Linkboost & GPU eX.
Would be nice if Nvidia could/would do software voltage controls like ATI already has.
If you wrote a nice letter to Nvidia saying that you liked ATI software voltage adjustments but you'd rather buy Nvidia cards if they had such feature, so maybe they would let you know their secret :D
I think it is pci-e clock frequency related.
i recall nvidia saying that their pci-e slots will run slightly faster.
If you take a look at guess2098's results of post 6
975x = 1.55v
680i = 1.64v
now stock pci-e is 100mhz
100 / 1.55 x 1.64 = 105.8 -> 106mhz
that to me seems easily possible. so maybe someone with a 8800 series card can check this out. i am thinking 120mhz on the pci-e shouldn't be too difficult, and if not, then this could be the easy way to voltmod your 8 series card :D
- Paul
so you're suggesting that upping the PCIE frequency results in a higher vgpu?
I've tried 120Mhz but that corrupted my SATA. I'll try with an IDE HD tonight.
My PCIE is currently running @ 115 MHz. No changes in VGPU/VMEM.