7900GT Bios Mod

Printable View

05-02-2006, 10:54 AM
DeToNaToR.cl

7900GT Bios Mod

:)

Found this on Nibitor, maybe it can work...

http://img366.imageshack.us/img366/4273/12v3hz.jpg

http://img366.imageshack.us/img366/3...eeditor3zd.jpg

http://img436.imageshack.us/img436/8...itor12v4ir.jpg

http://img436.imageshack.us/img436/9...itor15v2ib.jpg

http://img517.imageshack.us/img517/7...itor18v9rq.jpg

http://img366.imageshack.us/img366/4651/14v4yc.jpg

Can this be made?
05-02-2006, 06:11 PM
wizboy11

No, because of the over-voltage protection.

People have already tried that (i think).
05-03-2006, 12:01 AM
pocketbikeuk

does that protection only kick in at 1.55v i may be wrong but it does with my hard mod on my 7900gt manli
05-04-2006, 05:31 PM
KrauseOasis

I can change the voltage for 1.4 for biosmod? or so functions way mod with conductive ink?
05-04-2006, 06:00 PM
wizboy11

Quote:

Originally Posted by KrauseOasis

I can change the voltage for 1.4 for biosmod? or so functions way mod with conductive ink?

How many times must I say this (ok, this is the first time here, but I've said it else were), the 7900GT has overvoltage protection. It's not like the X1900 were you can just raise the voltage via ATI Tool.

You HAVE to make some sort of physical modifictaion to the card that somehow bypasses the overvoltage protection (like the Contuctive Ink mod, and the penicil vmem mod)
05-04-2006, 06:14 PM
KrauseOasis

Quote:

Originally Posted by wizboy11

How many times must I say this (ok, this is the first time here, but I've said it else were), the 7900GT has overvoltage protection. It's not like the X1900 were you can just raise the voltage via ATI Tool.

You HAVE to make some sort of physical modifictaion to the card that somehow bypasses the overvoltage protection (like the Contuctive Ink mod, and the penicil vmem mod)

Oks, thank you.
05-09-2006, 09:43 PM
Dave338

But OVP acts at about 1,28v ...

CAn be possible to raise vgpu to this limit with BIOS?? 0,05v maybe help to obtain 20-30mhz

salu2.
05-10-2006, 08:58 AM
frostedflakes

The problem is you're just changing the label for the VID code, which will do nothing. To change voltage in NiBiTor, you need to change the VID code. I've played around with my XFX 7900GT BIOS, and there is an extra VID code programmed in (00), but selecting it does nothing, so either NiBiTor isn't reading something right or the VID code is there but no hardware is on the card to support it. When you think about it, the latter makes sense, as the 7900GT is 1.2V default 2D and 3D. It'd just be a waste of money to include hardware for it to run at higher or lower voltages.
05-12-2006, 08:22 AM
Richteralan

Quote:

Originally Posted by wizboy11

How many times must I say this (ok, this is the first time here, but I've said it else were), the 7900GT has overvoltage protection. It's not like the X1900 were you can just raise the voltage via ATI Tool.

You HAVE to make some sort of physical modifictaion to the card that somehow bypasses the overvoltage protection (like the Contuctive Ink mod, and the penicil vmem mod)

While I know 7900GTs have OVP, but then why eVGA's beta 7900GT Bios shows a 1.35v core and some people reported that they can oc higher with the new beta bios?
Also eVGA's 7900GT Signature Series is using reference PCB also, with a 1.35V core.
05-12-2006, 04:12 PM
Roger_D25

Hey Guys,
I just recently got a BIOS from Mario who is an eVGA tech who I met in the eVGA forums. He made a modified BIOS to help with some of the issues we have had with the 7900GT's. I downloaded this BIOS version from him and just opened it with the new version of NiBiTor, version 2.9. Well unlike every other eVGA 7900GT BIOS that I've tried, this one reads to have a voltage of 1.35V. This wasn't some value that I changed with NiBiTor, it just reads that way on the VID tab.

The card I currnelty have is my second replacment and it's unfortunelty suffering from the same exact issues my original card did! Not only that but just like my first card, this one worked perfectly fine for 2 days and only then did it start artifacting, freezing, blinking on/off, ect... Anyway I just got my second RMA approved and my Cross-Shipment replacment card is on it's way so I don't want to try this BIOS (I'll have to ship this card back right after I get my new replacment card so I don't want to mess with it).

Anyway I'm just wondering if anyone has tried it and if so, did it actually change the voltage or not? From what I've just read above I'm quite sure it didn't or won't but it might be worth a shot? If anyone wants the BIOS just let me know and I'll e-mail it to you ASAP becasue I'd like to know if it will do anything?
05-12-2006, 04:36 PM
alexio

See: http://www.xtremesystems.org/forums/...ad.php?t=98911

Quote:

Originally Posted by alexio

Ok thank you. I have been looking at the bios-file and I'm fairly sure that this 1.35v card must have a different PCB from normal GT's, or the bios is just a big joke. I'm quite sure normal PCB's can only supply 1.1v and 1.2v. There are 2 VIDs and although you can name them to whatever you like it is impossible to change the voltage these VIDs give.

I hope I'm wrong though, ;)

Someone please test this bios.

Quote:

Originally Posted by Aleman

I tried this BIOS (from Mario Gastelum Tech at eVGA) but the volts did not change since I

1. could not OC any higher
2. the temperature did not change at all vs 1.2V

and the flash was successfull since I made a dump of the flashed bios and its the same as the one that supposes to be in the BIOS.

And the asnwer to the question what card he has he answered:

Quote:

Originally Posted by Aleman

eVGA 7900GT CO N563
05-12-2006, 06:33 PM
Richteralan

Quote:

Originally Posted by alexio

See: http://www.xtremesystems.org/forums/...ad.php?t=98911

And the asnwer to the question what card he has he answered:

Well then this rises a question.
If the card is physically limited to 1.2V
then why eVGA engineers put a 1.35V voltage call in BIOS?

And I'm curious about Signature Series' difference, too. It's been heard that the GPU is running 1.45V?
05-14-2006, 09:56 AM
nobody_here

i want to find out the low down on this

we have the GT, GT CO and GT CO SC cards which can be flashed to that new EVGA bios that supplies 1.35v to the core, so either the core is not getting 1.35v or there is no 1.2Xv voltage limit on a hardware level because many people have flashed their cards with this new BIOS and have reached significantly higher overclocks with it, so it does seem to allow 1.35v

then there's the Signature Series card that as Richteralan pointed out on the EVGA forums, at least according to Newegg pics, has several additional components on the back of the card, but i have the SS BIOS and it does indeed push 1.45v to the core

it runs 450/1320 at 2D level and 600/1600 at 3D level, runs 1.2v at 2D and 1.45v at 3D

memory timings are as follows

trc 12
trfc 14
tras 8
trp 4
trcdrd 4
trccdwr 4
trrd 2

if anyone can provide a high resolution pic of the back of a SS card, that would be great, will provide one of my Superclocked card this evening to compare

http://www.evga.com/community/messag...TOPIC_ID=14988
05-23-2006, 05:32 AM
Zyprex^^

Does the eVGA Signature Series' have the same layout as the other 7900GT cards?
05-24-2006, 06:01 PM
xenium

Ok, so what if you removed the two resistors that enable overcurrent protection and do the custom bios update??? Like as this pic shows, removing a coulple resistors disables overcurrent protection.

http://img117.imageshack.us/img117/5...tmod3ev.th.jpg

Then again, if you take the time to remove the two resistors you may as well voltmod while you at it. Just my 2 cents :cool:
05-24-2006, 08:22 PM
quijonsith

Read the sticky in this section. Those aren't overcurrent protection.

Quote:

These resistors. measuring 1,8Kohm, are meant to balance the current between the two pair of switching Mosfets to carry an equal amount of load.
05-29-2006, 05:50 AM
t024484

Quote:

Originally Posted by nobody_here

i want to find out the low down on this

we have the GT, GT CO and GT CO SC cards which can be flashed to that new EVGA bios that supplies 1.35v to the core, so either the core is not getting 1.35v or there is no 1.2Xv voltage limit on a hardware level because many people have flashed their cards with this new BIOS and have reached significantly higher overclocks with it, so it does seem to allow 1.35v

then there's the Signature Series card that as Richteralan pointed out on the EVGA forums, at least according to Newegg pics, has several additional components on the back of the card, but i have the SS BIOS and it does indeed push 1.45v to the core

it runs 450/1320 at 2D level and 600/1600 at 3D level, runs 1.2v at 2D and 1.45v at 3D

memory timings are as follows

trc 12
trfc 14
tras 8
trp 4
trcdrd 4
trccdwr 4
trrd 2

if anyone can provide a high resolution pic of the back of a SS card, that would be great, will provide one of my Superclocked card this evening to compare

http://www.evga.com/community/messag...TOPIC_ID=14988

To change the Vcore through Bios will only be possible with the semiconductor switches onboard in the area where the voltmodding is done.
There are very few boards that are equiped with those switches.

If they are not on your board, you can do whatever you like in your Bios, but it will have no effect on the Vcore.
If the EVGA Bios is giving a better score, it can only mean that something else is causing this, but not the Vcore.

I can hardly believe that you have selected the right timingset. It should be:31/39/22/9/10/6/8.
This reflects exactly the timing that Samsung specifies for the BC14.
05-29-2006, 07:40 AM
nobody_here

Quote:

Originally Posted by t024484

To change the Vcore through Bios will only be possible with the semiconductor switches onboard in the area where the voltmodding is done.
There are very few boards that are equiped with those switches.

If they are not on your board, you can do whatever you like in your Bios, but it will have no effect on the Vcore.
If the EVGA Bios is giving a better score, it can only mean that something else is causing this, but not the Vcore.

I can hardly believe that you have selected the right timingset. It should be:31/39/22/9/10/6/8.
This reflects exactly the timing that Samsung specifies for the BC14.

let me be clear....i did not flash the SS BIOS to my card

i'm just telling you what was in the BIOS that was extracted from a Signature Series card

trc 12
trfc 14
tras 8
trp 4
trcdrd 4
trccdwr 4
trrd 2

i didnt change them

of course the high vcore doesn't directly change performance, but the only way the Signatuire Series card can hit 700Mhz is because it runs 1.45v stock, whereas most 7900GT's run 1.2v stock

and yeah, i figured out that the BIOS changes do nothing, hardmod was required ;)
05-29-2006, 10:59 AM
wizboy11

Quote:

Originally Posted by nobody_here

let me be clear....i did not flash the SS BIOS to my card

i'm just telling you what was in the BIOS that was extracted from a Signature Series card

trc 12
trfc 14
tras 8
trp 4
trcdrd 4
trccdwr 4
trrd 2

i didnt change them

of course the high vcore doesn't directly change performance, but the only way the Signatuire Series card can hit 700Mhz is because it runs 1.45v stock, whereas most 7900GT's run 1.2v stock

and yeah, i figured out that the BIOS changes do nothing, hardmod was required ;)

It's pretty easy to do 1.45 with a hardmod. I'm running 1.4 and I'm hitting 680 on the core. Maby 1.45 will get me too 700mhz.
06-06-2006, 01:19 AM
Damien

Quote:

Originally Posted by wizboy11

It's pretty easy to do 1.45 with a hardmod. I'm running 1.4 and I'm hitting 680 on the core. Maby 1.45 will get me too 700mhz.

But what about those who have no experience with soldering? I have a CO coming tomorrow and i want to raise the volts :P