Author Topic: [MOVED] Hacking NVidia Cards into their Professional Counterparts  (Read 1644946 times)

0 Members and 2 Guests are viewing this topic.

Offline gordan

  • Frequent Contributor
  • **
  • Posts: 277
  • Country: 00
    • Tech Articles
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #425 on: June 28, 2013, 04:33:31 pm »
Ah, so you change the straps from the command line with nvflash.

I actually edit the straps while I'm editing the ROM. I find it easier because all I have to do afterwards is flash the ROM, although I do make sure that nvflash verifies my ROM first.

I may be wrong but IIRC NiBiTor doesn't update the strap checksum, only the full checksum.

I got to that point without heating up my soldering iron.
For me, Windows Quadro driver would not install if I do not change hard-straps. Matter a fact in some cases I could not even get to Windows GUI to install the driver.

Sounds strange, I've never seen an issue like that on any of the cards I modified, it always "just works".
Then again, modifying a GTX580 is a complete waste of time anyway. GTX480 is just as fast (in some cases faster due to dual DMA channels when modded to Quadro/Tesla), and is trivial to modify into a Q6000. I only modified my 580 because I already had it and I was seeing odd driver clashing when using a GeForce and a Quadro in the same system under Windows. One driver would end up driving both cards, usually the later one (the GeForce one has a higher version number).

Hmm... I wonder if this could be stripped out to reduce the BIOS to the old, EFI-less state that is more open to modifying. It'd also make it a lot more similar to the previous BIOSes in terms of understanding what the various bits do.
That should be doable, but you have to make sure that you include all ROM parts (minus the EFI). In Fermi cards, the ROM had two parts (one was the vbios and the other another device, I believe HDMI audio).

Indeed, but you get the same thing on Kepler, they have a HDMI audio device, too.

I don't own any EFI motherboards, but I would have thought the whole EFI wrapping should be strippable out from the VBIOS. Once you skip past the 0x400 bytes of header, the rest of the BIOS is similar in terms of offsets of known areas to the Fermi BIOSes.
IIRC, according to the PCI firmware spec the wrapping stuff ends up being ignored on older computers with BIOS, until the 55 AA signature is detected. On the new UEFI, they will probable read the wrapper and then read the rest (55 AA, on).

Handy, so you could effectively s/.*55AA// and strip out the EFI capability and defeat crypto. Nice. Presumably the trailing garbage would just get ignored then.

The only question is whether there is an extra ID bit in Kepler soft straps. These are not yet fully documented. It could be one of the unknown bits (but it's not the unknowns next to ID bit 4, I tried those). The reason I say that is because until Kepler, all cards were modifiable using only soft-straps into any other card sporting the same GPU and memory type. And I only mention memory type because the GDDR3 GTS450 differed in more than just the last 5 bits of device ID, I had not seen such a case before.

What I'll probably do on my GTX680 is mod the resistor slot controlling the 3rd nibble by adding a dip switch, so I can toggle that bit between 1 and 0. From there on, I can modify the GTX680 into any card sporting the GK104 chip using the switch the 5 soft-strapped bits.
« Last Edit: June 28, 2013, 04:36:05 pm by gordan »
 

Offline amigo

  • Regular Contributor
  • *
  • Posts: 108
  • Professional Wannabe
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #426 on: June 28, 2013, 06:21:48 pm »
I may be wrong but IIRC NiBiTor doesn't update the strap checksum, only the full checksum.
Who said anything about NiBiTor? :)

I edit everything by hand in a hex editor - I've done so many things now that I know where to look and what to do with it blindfolded.

Sounds strange, I've never seen an issue like that on any of the cards I modified, it always "just works".
Then again, modifying a GTX580 is a complete waste of time anyway. GTX480 is just as fast (in some cases faster due to dual DMA channels when modded to Quadro/Tesla), and is trivial to modify into a Q6000. I only modified my 580 because I already had it and I was seeing odd driver clashing when using a GeForce and a Quadro in the same system under Windows. One driver would end up driving both cards, usually the later one (the GeForce one has a higher version number).
I deliberately chose GTX 570 because GTX 4xx are running much hotter than 5xx and if they even have any performance improvements over 5xx I think it's negligible. Altough even GTX 5xx run hot, heck they are all crap due to the terrible NVidia design which makes the GPU run additional 10-15C hotter if you plug two monitors into the card.

Handy, so you could effectively s/.*55AA// and strip out the EFI capability and defeat crypto. Nice. Presumably the trailing garbage would just get ignored then.

The only question is whether there is an extra ID bit in Kepler soft straps. These are not yet fully documented. It could be one of the unknown bits (but it's not the unknowns next to ID bit 4, I tried those). The reason I say that is because until Kepler, all cards were modifiable using only soft-straps into any other card sporting the same GPU and memory type. And I only mention memory type because the GDDR3 GTS450 differed in more than just the last 5 bits of device ID, I had not seen such a case before.
You can't defeat the signature but I think Linux will ignore it altogether so it does not matter if it's there. The only reason it would matter is to Windows.

Also, because of all the new crap added to the ROM the EEPROM is bigger too- instead of 256KB it's 512KB now thus requires soldering when modding some cards.
 

Offline gordan

  • Frequent Contributor
  • **
  • Posts: 277
  • Country: 00
    • Tech Articles
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #427 on: June 28, 2013, 06:47:36 pm »
I may be wrong but IIRC NiBiTor doesn't update the strap checksum, only the full checksum.
Who said anything about NiBiTor? :)

I edit everything by hand in a hex editor - I've done so many things now that I know where to look and what to do with it blindfolded.

Can you please share the strap checksum algorithm? How do you compute it after modifying the strap manually?

Sounds strange, I've never seen an issue like that on any of the cards I modified, it always "just works".
Then again, modifying a GTX580 is a complete waste of time anyway. GTX480 is just as fast (in some cases faster due to dual DMA channels when modded to Quadro/Tesla), and is trivial to modify into a Q6000. I only modified my 580 because I already had it and I was seeing odd driver clashing when using a GeForce and a Quadro in the same system under Windows. One driver would end up driving both cards, usually the later one (the GeForce one has a higher version number).
I deliberately chose GTX 570 because GTX 4xx are running much hotter than 5xx and if they even have any performance improvements over 5xx I think it's negligible. Altough even GTX 5xx run hot, heck they are all crap due to the terrible NVidia design which makes the GPU run additional 10-15C hotter if you plug two monitors into the card.

Most of the reason why 580 runs cooler than the 480 is the power management that throttles certain operations. That plus they cut out the 2nd DMA engine. GF100 and GF110 are nearly identical silicon-wise.

Handy, so you could effectively s/.*55AA// and strip out the EFI capability and defeat crypto. Nice. Presumably the trailing garbage would just get ignored then.

The only question is whether there is an extra ID bit in Kepler soft straps. These are not yet fully documented. It could be one of the unknown bits (but it's not the unknowns next to ID bit 4, I tried those). The reason I say that is because until Kepler, all cards were modifiable using only soft-straps into any other card sporting the same GPU and memory type. And I only mention memory type because the GDDR3 GTS450 differed in more than just the last 5 bits of device ID, I had not seen such a case before.
You can't defeat the signature but I think Linux will ignore it altogether so it does not matter if it's there. The only reason it would matter is to Windows.

Also, because of all the new crap added to the ROM the EEPROM is bigger too- instead of 256KB it's 512KB now thus requires soldering when modding some cards.

I presume you speak of the 7xx series - My GTX680 ROM is 180KB.
As for Windows - the only reason for using that is disappearing with Steam acquiring Linux support.
« Last Edit: June 28, 2013, 10:53:41 pm by gordan »
 

Offline amigo

  • Regular Contributor
  • *
  • Posts: 108
  • Professional Wannabe
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #428 on: June 29, 2013, 01:14:15 am »
Can you please share the strap checksum algorithm? How do you compute it after modifying the strap manually?
The algorithm is the same as for the ROM itself: 0x100 - (S & 0xFF), where S is sum of bytes from offset 0x58 to 0x6a (0x6a is always 0xA5, it's a HWINFO signature) in a standard vbios image. Offset 0x6b is then the checksum itself which is not part of the S.

I presume you speak of the 7xx series - My GTX680 ROM is 180KB.
Actually GTX 680 ROM is over 200KB due to the EFI portion and all the other crap like secure signature and new wrappers, etc. That is on a GTX 680 2 or 4GB I've seen.

As for Windows - the only reason for using that is disappearing with Steam acquiring Linux support.
Hear, hear!
 

Offline gamezr2ez

  • Contributor
  • Posts: 30
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #429 on: June 29, 2013, 03:20:56 am »
Actually GTX 680 ROM is over 200KB due to the EFI portion and all the other crap like secure signature and new wrappers, etc. That is on a GTX 680 2 or 4GB I've seen.

My Asus is 180k.

A GV-N680OC-2GD is 64k.

Of note, the gigabyte one does not have uefi gop support; no space I am guessing.
 

Offline gordan

  • Frequent Contributor
  • **
  • Posts: 277
  • Country: 00
    • Tech Articles
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #430 on: June 29, 2013, 12:11:36 pm »
@amigo
Thanks for the checksum algorithm.

@gamezr2ez
Thanks for the info regarding the GB card.

I might just grab a copy of that BIOS from the TPU database and see if I can work out which bits can thus be stripped out from the Asus BIOS to remove all the UEFI crap. I suspect that with the UEFI headers and the other stuff removed, the BIOS will in fact be parseable by all the old tools that work on the previous generations' BIOSes. The only issue is trying to acquire a cheap sacrificial 680 for the research (I don't want to brick my shiny 4GB Gainward Phantom).
 

Offline gamezr2ez

  • Contributor
  • Posts: 30
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #431 on: June 30, 2013, 02:24:44 am »
As I am aware there is no UEFI stuff in the ASUS bios by default. I believe they have released an unofficial-official hybrid bios. By default though that is all left out.

Side thought, why would you care about stripping out anything from the BIOS; it wont result in added performance.
 

Offline amigo

  • Regular Contributor
  • *
  • Posts: 108
  • Professional Wannabe
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #432 on: June 30, 2013, 02:39:09 am »
As I am aware there is no UEFI stuff in the ASUS bios by default. I believe they have released an unofficial-official hybrid bios. By default though that is all left out.

Side thought, why would you care about stripping out anything from the BIOS; it wont result in added performance.

If I'm correct, current vbios editing tools do not support the new structures and so trying to edit those ROMs is not possible at the moment. Stripping the new wrappers exposes the old format which these editors can recognize and use.
 

Offline gordan

  • Frequent Contributor
  • **
  • Posts: 277
  • Country: 00
    • Tech Articles
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #433 on: June 30, 2013, 11:05:15 pm »
Has anybody found where the second set of device ID strap resistors is on the GTX690 yet? Or on the Titan?
 

Offline gamezr2ez

  • Contributor
  • Posts: 30
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #434 on: July 01, 2013, 12:02:43 am »
Has anybody found where the second set of device ID strap resistors is on the GTX690 yet?

Pretty sure gnif lost his card finding those. Check the first post again.
 

Offline gordan

  • Frequent Contributor
  • **
  • Posts: 277
  • Country: 00
    • Tech Articles
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #435 on: July 01, 2013, 06:27:06 am »
I did, but that hasn't been updated in months and I joined the thread later on. If there's still a funding gap and the research into the 2nd set of straps is ongoing, I'm happy to contribute but I haven't seen a post from gnif in the last few pages of the thread.
 

Offline gamezr2ez

  • Contributor
  • Posts: 30
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #436 on: July 01, 2013, 06:50:29 am »
As per everything in the first post:

GPU 0
GPU 1
 

Offline bdx

  • Newbie
  • Posts: 9
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #437 on: July 03, 2013, 06:00:15 am »
Does anyone have experience modding to the Tesla grid k2, the nvs grid k2, the geforce k2 rather than the generic k2?
How about the k340 or k520?
 :-//
 What are the differences in these cards?

http://browse.feedreader.com/c/LaptopVideo2Go_News/358076787

 

Offline gordan

  • Frequent Contributor
  • **
  • Posts: 277
  • Country: 00
    • Tech Articles
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #438 on: July 03, 2013, 06:51:23 am »
Where did you find references to these cards? I can't find them in the supported list for the latest driver:
http://us.download.nvidia.com/XFree86/Linux-x86/319.23/README/supportedchips.html
Which means they are unlikely to exist (yet, at least).
 

Offline Soulnight

  • Contributor
  • Posts: 28
Passiv 3D with dual projection and Nvidia 3D vision
« Reply #439 on: July 13, 2013, 03:25:48 pm »
Hello everybody,

I'm new on this forum and I would like to thank everybody (and especially gnif) for the mod of geforce GTX 680 to Quadro K5000.

I'm a gamer. I don't want more performance but just one (quadro) functionnality more.

I play games in 3D with the help of nvidia 3D vision on my monitor (input signal 1080P 3D 120Hz).
I would like to do the same but with passive 3D dual projection. I need therefore the option of a quadro K5000 to activate the possibility of passive stereo with 2 projectors connected to the graphic card. Here is the link to the procedure to do passive stereo with a quadro.

http://nvidia.custhelp.com/app/answers/detail/a_id/3012/~/how-to-configure-passive-or-dual-pipe-stereo-with-quadro-cards-in-windows-7.

Could someone please confirm that the option is available with a modded GTX 680 to quadro K5000?

Thank you,
Soulnight  ;)
 

Offline myweb

  • Newbie
  • Posts: 9
  • Country: by
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #440 on: July 17, 2013, 06:09:04 pm »
Dear All,

What inexpensive card would you recommend to try the modifications? Is it correct, that the most inexpensive card is GT640?
 

Offline Jager

  • Contributor
  • Posts: 30
Re: Passiv 3D with dual projection and Nvidia 3D vision
« Reply #441 on: July 17, 2013, 09:41:24 pm »
Hello everybody,

I'm new on this forum and I would like to thank everybody (and especially gnif) for the mod of geforce GTX 680 to Quadro K5000.

I'm a gamer. I don't want more performance but just one (quadro) functionnality more.

I play games in 3D with the help of nvidia 3D vision on my monitor (input signal 1080P 3D 120Hz).
I would like to do the same but with passive 3D dual projection. I need therefore the option of a quadro K5000 to activate the possibility of passive stereo with 2 projectors connected to the graphic card. Here is the link to the procedure to do passive stereo with a quadro.

http://nvidia.custhelp.com/app/answers/detail/a_id/3012/~/how-to-configure-passive-or-dual-pipe-stereo-with-quadro-cards-in-windows-7.

Could someone please confirm that the option is available with a modded GTX 680 to quadro K5000?

Thank you,
Soulnight  ;)

I'm on same boat. I'm just trying to figure best way to get Dual projection setup with omega filters to work. There is couple methods to get frame synced dual projection to work. One is using Tridef SBS with AMD eyefinity 3840x1080 and second method is using Quadro. I want to use 3D Vision and i'm also ready to try to mod my 670 to K5000 to get quadro features. I think 3D Vision dual projection will work with modded card. I'm bit unsure however if quadro gives framesynced dual output from one card...
 

Offline gordan

  • Frequent Contributor
  • **
  • Posts: 277
  • Country: 00
    • Tech Articles
Re: Hacking NVidia Cards into their Professional Counterparts
« Reply #442 on: July 18, 2013, 06:15:33 pm »
@myweb - what is the reason for your looking to modify a card? Depending on the level of performance and features you require will dictate which series card you are best off with (e.g. 4xx series for Quadro x000 or 6xx series for Quadro Kx000 / Grid).
 

Offline Soulnight

  • Contributor
  • Posts: 28
Re: Passiv 3D with dual projection and Nvidia 3D vision
« Reply #443 on: July 20, 2013, 10:03:06 am »
Hello everybody,

I'm new on this forum and I would like to thank everybody (and especially gnif) for the mod of geforce GTX 680 to Quadro K5000.

I'm a gamer. I don't want more performance but just one (quadro) functionnality more.

I play games in 3D with the help of nvidia 3D vision on my monitor (input signal 1080P 3D 120Hz).
I would like to do the same but with passive 3D dual projection. I need therefore the option of a quadro K5000 to activate the possibility of passive stereo with 2 projectors connected to the graphic card. Here is the link to the procedure to do passive stereo with a quadro.

http://nvidia.custhelp.com/app/answers/detail/a_id/3012/~/how-to-configure-passive-or-dual-pipe-stereo-with-quadro-cards-in-windows-7.

Could someone please confirm that the option is available with a modded GTX 680 to quadro K5000?

Thank you,
Soulnight  ;)

I'm on same boat. I'm just trying to figure best way to get Dual projection setup with omega filters to work. There is couple methods to get frame synced dual projection to work. One is using Tridef SBS with AMD eyefinity 3840x1080 and second method is using Quadro. I want to use 3D Vision and i'm also ready to try to mod my 670 to K5000 to get quadro features. I think 3D Vision dual projection will work with modded card. I'm bit unsure however if quadro gives framesynced dual output from one card...

I am happy that I am not alone!  :clap:

I also want to use the omega filters for the dual projector setup. And I think I will buy two projectors ACER h9500 (ONLY 2D and costs just 850€ with lens shift and 1300 Lumens calibrated!). I'm just worried about the 1:1 Hdmi Mapping problem but that's another story...

I know the tridef solution but it's not ideal and doesn't work with all the games...and you MUST use tridef.
I would like to have a choice and to be able to play every games: therefore the quadro solution.
The real quadro K5000 is not good enough for 3d games (and is expensive) and I really think that a modded gtx 680 (or gtx 670) is the perfect solution to get the functionnality of the quadro AND the 3D games performances of the GTX 680!

I am pretty sure that a quadro card gives framesynced dual output from one card...
See the link: http://nvidia.custhelp.com/app/answers/detail/a_id/3012/~/how-to-configure-passive-or-dual-pipe-stereo-with-quadro-cards-in-windows-7.

The real question is: does it work with a modded gtx 680 as well?

Plus, I would like to go sli with 2 modded GTX 680 into quadro K5000. But I've read that the quadro just support sli for choosen "complete Workstation" from dell etc... However it may still be possible to do it since they won't be true K5000.

As someone succeded in activating SLI with 2 modded GTX 680 into quadro K5000?

@ Jager: how far are you from testing the dual projector setup with the modded gtx 670? When?  ::)

Thank you!
Soulnight  ;)



« Last Edit: July 20, 2013, 10:06:40 am by Soulnight »
 

Offline Jager

  • Contributor
  • Posts: 30
Re: Passiv 3D with dual projection and Nvidia 3D vision
« Reply #444 on: July 20, 2013, 12:40:52 pm »

I am happy that I am not alone!  :clap:

I also want to use the omega filters for the dual projector setup. And I think I will buy two projectors ACER h9500 (ONLY 2D and costs just 850€ with lens shift and 1300 Lumens calibrated!). I'm just worried about the 1:1 Hdmi Mapping problem but that's another story...

I know the tridef solution but it's not ideal and doesn't work with all the games...and you MUST use tridef.
I would like to have a choice and to be able to play every games: therefore the quadro solution.
The real quadro K5000 is not good enough for 3d games (and is expensive) and I really think that a modded gtx 680 (or gtx 670) is the perfect solution to get the functionnality of the quadro AND the 3D games performances of the GTX 680!

I am pretty sure that a quadro card gives framesynced dual output from one card...
See the link: http://nvidia.custhelp.com/app/answers/detail/a_id/3012/~/how-to-configure-passive-or-dual-pipe-stereo-with-quadro-cards-in-windows-7.

The real question is: does it work with a modded gtx 680 as well?

Plus, I would like to go sli with 2 modded GTX 680 into quadro K5000. But I've read that the quadro just support sli for choosen "complete Workstation" from dell etc... However it may still be possible to do it since they won't be true K5000.

As someone succeded in activating SLI with 2 modded GTX 680 into quadro K5000?

@ Jager: how far are you from testing the dual projector setup with the modded gtx 670? When?  ::)

Thank you!
Soulnight  ;)

Ordered resistors from ebay, those are quite tiny(1mm*0,5mm). I have done some researching and i think up to 2 GPU SLI is no problem to get synced frames, above that K5000 needs Sync card and it should work with modded ones too. After all surround setups for desktops are synced. When i get those resistors(39K is closest to 40K i found, hope it works) i do this mod immediately.
 

Offline Soulnight

  • Contributor
  • Posts: 28
Re: Passiv 3D with dual projection and Nvidia 3D vision
« Reply #445 on: July 20, 2013, 06:04:55 pm »
Ordered resistors from ebay, those are quite tiny(1mm*0,5mm). I have done some researching and i think up to 2 GPU SLI is no problem to get synced frames, above that K5000 needs Sync card and it should work with modded ones too. After all surround setups for desktops are synced. When i get those resistors(39K is closest to 40K i found, hope it works) i do this mod immediately.

I hope it will work with a 39K resistor... How do you know the size of the resistors you must take? What are the references?

The problem for SLI with quadro cards is that Nvidia doesn't enable the SLI function for quadro cards UNLESS they are used in specific Workstation that they have certified!

Here the link to nvidia sh**t:
http://www.nvidia.com/object/quadro_sli_compatible_systems.html

But I do hope that because they are "just" modded GTX 680 the false K5000 can still do sli WITHOUT working with a specific nvidia certified workstation...
« Last Edit: July 20, 2013, 06:08:43 pm by Soulnight »
 

Offline Jager

  • Contributor
  • Posts: 30
Re: Passiv 3D with dual projection and Nvidia 3D vision
« Reply #446 on: July 21, 2013, 06:28:10 am »
Ordered resistors from ebay, those are quite tiny(1mm*0,5mm). I have done some researching and i think up to 2 GPU SLI is no problem to get synced frames, above that K5000 needs Sync card and it should work with modded ones too. After all surround setups for desktops are synced. When i get those resistors(39K is closest to 40K i found, hope it works) i do this mod immediately.

I hope it will work with a 39K resistor... How do you know the size of the resistors you must take? What are the references?

The problem for SLI with quadro cards is that Nvidia doesn't enable the SLI function for quadro cards UNLESS they are used in specific Workstation that they have certified!

Here the link to nvidia sh**t:
http://www.nvidia.com/object/quadro_sli_compatible_systems.html

But I do hope that because they are "just" modded GTX 680 the false K5000 can still do sli WITHOUT working with a specific nvidia certified workstation...

Damn missed that one! This is bad news indeed :(. I hope it is fixable with some softmodding/just work with modded one. 690 should be affected as well. As for size of resistor, if mean physical size, i just measured those and other info comes from this thread.

edit: There is HyperSLI for non sli motherboards, maybe this works if needed...

edit2:Seems that HyperSLI will NOT work with Quadros.
« Last Edit: July 21, 2013, 08:41:27 am by Jager »
 

Offline Soulnight

  • Contributor
  • Posts: 28
Re: Passiv 3D with dual projection and Nvidia 3D vision
« Reply #447 on: July 21, 2013, 07:15:30 am »
Damn missed that one! This is bad news indeed :(. I hope it is fixable with some softmodding/just work with modded one. 690 should be affected as well. As for size of resistor, if mean physical size, i just measured those and other info comes from this thread.

Yeah I mean the physical size of the resistors to use. Could someone post a link to "right" resistors on the internet? Thank you...
 

Offline Jager

  • Contributor
  • Posts: 30
Re: Passiv 3D with dual projection and Nvidia 3D vision
« Reply #448 on: July 21, 2013, 07:33:56 am »
Damn missed that one! This is bad news indeed :(. I hope it is fixable with some softmodding/just work with modded one. 690 should be affected as well. As for size of resistor, if mean physical size, i just measured those and other info comes from this thread.

Yeah I mean the physical size of the resistors to use. Could someone post a link to "right" resistors on the internet? Thank you...

I ordered this set. Few extra ones i think :) >> http://www.ebay.com/itm/0402-SMD-SMT-Chip-Resistor-Assortment-Kit-170-value-x-20-pcs-component-pack-/200940135046?pt=LH_DefaultDomain_0&hash=item2ec8f72286

edit:Wrong link
« Last Edit: July 26, 2013, 05:56:31 am by Jager »
 

Offline gordan

  • Frequent Contributor
  • **
  • Posts: 277
  • Country: 00
    • Tech Articles
Re: Passiv 3D with dual projection and Nvidia 3D vision
« Reply #449 on: July 21, 2013, 10:42:30 am »
The real quadro K5000 is not good enough for 3d games (and is expensive) and I really think that a modded gtx 680 (or gtx 670) is the perfect solution to get the functionnality of the quadro AND the 3D games performances of the GTX 680!

How exactly do you figure a K5000 isn't good enough when a GTX680 is? The spec between them is near identical.

I am pretty sure that a quadro card gives framesynced dual output from one card...

I'm pretty sure all Nvidia cards do, going back at least to 8xxx series (I am running an IBM T221 DG1 off an 8800GT with two DVI outputs, and that monitor supposedly requires genlocked inputs).
Vsync is a subtly different issue, and again, I'm pretty sure all Nvidia cards do run vsynced across separate outputs if they are told to run multiple screens as a same frame buffer. Again, on a T221 DG1 ATI cards (4850, cannot use 5xxx+ since they only comne with single DL-DVI outputs) produce tearing along the middle of the screen (genlocked but not vsynced), but Nvidia cards (tried with 8800GT, various 4xx, 580 and 680 cards, quadrified and vanilla, and a Quadro 2000) do not produce the tearing artifact - which implies they all run multiple screens vsynced. So provided you configure your setup correctly, it should work just fine.

Plus, I would like to go sli with 2 modded GTX 680 into quadro K5000. But I've read that the quadro just support sli for choosen "complete Workstation" from dell etc... However it may still be possible to do it since they won't be true K5000.

I didn't think Quadro cards came with SLI bridge ports. My Quadro 2000 certainly doesn't.

One thing is for sure, though, two separate cards won't be providing genlocked output, which means they won't be vsynced either.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf