Well sometimes even the most obvious things can get overlooked, even in an enviroment that has numerous tests and such. Maybe the post september 2007 mac pros weren't considered any different by those responsible for testing as there had never been any conflicting issues between them.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Upgrading old Mac Pro to 8800 GT graphics?
- Thread starter Fezzy
- Start date
- Sort by reaction score
Well sometimes even the most obvious things can get overlooked, even in an enviroment that has numerous tests and such. Maybe the post september 2007 mac pros weren't considered any different by those responsible for testing as there had never been any conflicting issues between them.
I see. I doubt it would work, to my knowledge no Mobility ROM has ever been flashed into a desktop card successfully. They are likely quite different animals.
As flashes go, the current crop of PCI-E cards are pretty low-risk...but you'd probably want to have a backup card available for troubleshooting.
I wonder what ROM they used over at netkas.org then for their 3850/3870.
I have a GMA 3100 on my motherboard so I won't have to worry that much.
My MacBook serves me well and so does my minitower.
Oh wow.. I almost put BUY NOW at the ibuypower computer store.
Well gee this is VERY enlightening news. It still could be some workaround "change of mind" because of all the complaints, but that sounds pretty genuine to me.
I suppose I could wait til next month, February or when does macworld come? If not by Macworld a fix comes to place I can make my decision then.
I still keep my words and I don't think i jumped to any irrational behavior, because the way it seemed was that our computers were left without any options. The damage is already done and some people have already left.
You guys do realize next month the 9xxx series Nvidia cards will be out? But then is it possible to write an efi for that ?
If this is true, I will give Apple benefit of the doubt and it put it behind me. But I always be on guard and view them as a corporation.
I think I'm more neutral to mac than ever now.
Thank you again, ANtzona, you saved a lot of our hopes!
Well gee this is VERY enlightening news. It still could be some workaround "change of mind" because of all the complaints, but that sounds pretty genuine to me.
I suppose I could wait til next month, February or when does macworld come? If not by Macworld a fix comes to place I can make my decision then.
I still keep my words and I don't think i jumped to any irrational behavior, because the way it seemed was that our computers were left without any options. The damage is already done and some people have already left.
You guys do realize next month the 9xxx series Nvidia cards will be out? But then is it possible to write an efi for that ?
If this is true, I will give Apple benefit of the doubt and it put it behind me. But I always be on guard and view them as a corporation.
I think I'm more neutral to mac than ever now.
Thank you again, ANtzona, you saved a lot of our hopes!
Sigh...I'm not exactly surprised that all this is happening.
Apple's big problem here is that it's afraid of releasing specs and code to anyone without being over their constant shoulder about it. They could have not had this major issue with the 8800GT if they released proper specs to nVidia, since nVidia seems well aware of the PCIe issue prior to release. Hell, if they gave the OSX Dev kits to these people, they could actually develop ForceWare drivers compatible with OSX, and then you could put in any nVidia card you damn please into the Mac Pro. And we wouldn't be talking about this as much as we are. (Theoretically, the same argument applies to AMD and their Catalyst drivers)
I know my first post comes off as trollish, but being in the PC graphics/gaming business, one has to be open, because otherwise you lose money. Apple is doing the exact opposite.
Apple's big problem here is that it's afraid of releasing specs and code to anyone without being over their constant shoulder about it. They could have not had this major issue with the 8800GT if they released proper specs to nVidia, since nVidia seems well aware of the PCIe issue prior to release. Hell, if they gave the OSX Dev kits to these people, they could actually develop ForceWare drivers compatible with OSX, and then you could put in any nVidia card you damn please into the Mac Pro. And we wouldn't be talking about this as much as we are. (Theoretically, the same argument applies to AMD and their Catalyst drivers)
I know my first post comes off as trollish, but being in the PC graphics/gaming business, one has to be open, because otherwise you lose money. Apple is doing the exact opposite.
It occurs to me that the reason that this has happened is that Apple outsourced development of the new graphics cards, together with firmware and software to the respective manufacturers. They used both for competitive advantage. They supplied pre-production Mac pros to both manufacturers, and they came up with the production 2600 and the 8800, complete with firmware and software. ATI did a better job. Nvidia dropped the ball and didn't make sure there was compatibility with previous logic boards.
This would explain everything, including Apple's "deer in the headlights" response.
This would explain everything, including Apple's "deer in the headlights" response.
Actually, as I understand it, it seems like ATI used stock 2007 Mac Pros to design the card on.
Perhaps only NVIDIA got pre-production units?
Regardless, I touched on this earlier and it seems fully plausible.
As you understand it? Does that mean you have inside knowledge?
To me it's completely implausible that Apple would supply pre-production units only to one supplier, especially if the other supplier was tasked with making the "upgrade" card.
If "touched on this earlier" means talking about reference cards, I don't see the relevance.
I decided to take a look at the Apple documentation on the issue.. and although I can see why some might be confused and claim certain things and let the dog loose.. I don't believe the article says cards are incompatible between Mac Pros, but rather between Mac Pros and G5s..
http://docs.info.apple.com/article.html?artnum=305346
The way it's written, might make you think the three are different, but it's simply stating which cards come with what, but for example if you look at the 4500 Quadro, you notice the part works for both models, just as I imagine the 1900 and 7300 work for the new Mac Pro. In that case, and if you keep reading the article as it was written, it doesn't make sense to say the "early 2008 cards" would not work with the pre-2008 MPs..
Hope I made any sense, but I'm a lot less worried now.
http://docs.info.apple.com/article.html?artnum=305346
The way it's written, might make you think the three are different, but it's simply stating which cards come with what, but for example if you look at the 4500 Quadro, you notice the part works for both models, just as I imagine the 1900 and 7300 work for the new Mac Pro. In that case, and if you keep reading the article as it was written, it doesn't make sense to say the "early 2008 cards" would not work with the pre-2008 MPs..
Hope I made any sense, but I'm a lot less worried now.
I don't think it's a case of giving dev kits to nv and ati more a case of them WANTING to develop for OS X. I dare say Apple have to pay nv and ATi to support them with their low (but on the increase!) market share.
Anyway isn't a standard Mac Pro a "dev kit" for OS X?
lol it's just too funny that people running hackintoshs can get cards working with just a bit of hacking while us the Mac faithful have to wait for "Apple approved" cards to come out only to find they may not work.
The hackintosh community is putting Apple to shame, tsk tsk!
The hackintosh community is putting Apple to shame, tsk tsk!
Someone from ATI has been quoted as saying they used 2007 retail Mac Pros developing the firmware for the cards. It was on one of the news sites, I'm searching for the source now.
I agree that it seems implausible. It also seems implausible to me that no one bothered to test the 8800s on old Mac Pros, so there you go.
"Touched on it earlier" means I mentioned that ATI obviously did a better job designing firmware for this batch of cards several pages back.
I was agreeing with you, don't get so defensive. We're all on the same side here.
The absolute worst part about it is that Apple could let us boot PC versions of these cards if they just wrote an EFI module to hook VGA BIOS on boot... In fact, you know, that would be really easy to do?
EFI extensions are written in C. Hell guys, I think I just had a REALLY good idea. Just let me go download some documentation from Intel...
As a long time Mac user, I know three people attempting "hackintoshing".
Their results could not by any means be construed as equivalent to the usual Mac experience. Nice hobby, but not a viable method for anyone that values stability, or updatability.
If you manage to get PC cards working in OS X I may just have to betroth my (unborn, in fact unconceived as of yet[EDIT]but I'm working on it with as many people as I can[/EDIT]) daughter to you and worship you as a god of OS X and the saviour of Apple for the rest of my life.
Sorry to be coming across as defensive, just wanted clarification. Don't bother searching, I believe you.
You're right about ATI doing a better job.
Ain't that the truth.
Somebody should pop this in a mail to Steve. It's almost embarrasing.
Somebody should pop this in a mail to Steve. It's almost embarrasing.
After a bit more research, this might be easier than I thought it would be, yet harder at the same time.
I'd write a "fake" UGA driver that is loaded by EFI at boot time. This "fake" UGA driver would report that a video card was present, clock rate, frame buffer, address space, etc, etc.
This would make EFI think there was a valid video card present and proceed. The thing is, there *is* a valid video card present, it just doesn't have UGA firmware of its own. The information that my fake driver reports would be correct, so when the NVIDIA or ATI kexts load up, they'll take care of everything else.
Now, the big issue with this is as follows:
I'd need to write a different fake UGA driver for every variation of PC graphics card. By that I mean, an eVGA brand 8800GT would require a different UGA driver than, say, a PNY brand 8800GT.
Is there an easier way to do this? Yes, I think there may be.
People in the Hackintosh community have something called Titan.kext.
What Titan does is manually set all the video card information in the IOReg plane at boot time, at the operating system level, rather than the EFI level.
The problem with just outright using something like Titan is that your Mac Pro is not going to boot if you just have a PC video card installed.
EFI won't be able to initialize a video device and thus will refuse to boot.
There's two ways around this:
1) Create a single fake UGA driver that tells EFI there's a valid video card in the system, thus allowing the system to boot. (You won't have any graphics on screen from power up until the OS loads with this method.)
2) Keep a 7300GT or HD 2600 installed in the system so it has a valid video device on boot.
Either way, once the system gets into the OS loading stage and Titan is loaded, we'd dump the appropriate values for your video card right into the IOReg and boom, video fires up off your shiny PC 8800GT (or whatever other compatible PC card you have).
Now, I don't think Titan has been updated to work with Leopard, but his website has all the basic information available to make my own version, specifically for Mac Pro users that works with 10.5.
This sounds complicated, but once it gets programed and tested, it should be really simple for an end user to install. The best part is, it won't require any flashing!
Wow timb, if you could get that working (even just partially [working with one type of card])...
Especially because my x1900 just died. I don't want to have to spend $400 on a card that has previously given me troubles.
Especially because my x1900 just died. I don't want to have to spend $400 on a card that has previously given me troubles.
hell even if we are limited to only 1 pc graphics card version thats still better than nothing. yay!
No promises guys, but I'll give it my best shot.
Working on programming a few test kexts now. I'll probably end up ordering a PC 8800GT this weekend. I'd imagine 10.5.2 will be out on Monday, but if not, that's what I'm an ADC member for.
Who are you referring to exactly?
Working on programming a few test kexts now. I'll probably end up ordering a PC 8800GT this weekend. I'd imagine 10.5.2 will be out on Monday, but if not, that's what I'm an ADC member for.
Who are you referring to exactly?
No promises guys, but I'll give it my best shot.
Working on programming a few test kexts now. I'll probably end up ordering a PC 8800GT this weekend. I'd imagine 10.5.2 will be out on Monday, but if not, that's what I'm an ADC member for.
Who are you referring to exactly?
No offense timb, and good luck with it, but it seems to me that it's pretty obvious from previous posts that we will have an official solution to this problem sooner or later, that will be actually supported by Apple.
And that support is less likely to leave users out in the cold.
Cheers!
Official Apple solutions are well and fine, but my solution has the benefit of allowing a much wider (and cheaper) range of hardware. Want to use a X1950? No problem. 7800GT? No problem.
I had actually intended to research and follow through with this idea in late 2006 (when Titan first came to be), however I never really got around to it.
You're welcome to sit around and wait for Apple if you'd like, but I for one am tired of it. I'm going to be proactive.
I had actually intended to research and follow through with this idea in late 2006 (when Titan first came to be), however I never really got around to it.
You're welcome to sit around and wait for Apple if you'd like, but I for one am tired of it. I'm going to be proactive.