Not to create an outburst, but the dearth of or complete lack of professional video hardware for the Mac has always been frustrating to deal with. The Quadro has been selling well on the Mac as far as I can tell, so why is there no FirePro development yet? Has anyone heard of any development to get the FirePro on OSX natively? I remember years before any card manufacturers officially supported Linux there were companies like Xi Graphics that were writing Linux drivers for the highest end Nvidia and ATi workstation cards and selling the drivers to a lot of different studios. I wish there was atleast one hardcore coding group on the Mac side that targeted writing drivers for these highend video cards to run on the Mac.
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.
Where's the freakin Mac FirePro?
- Thread starter Imhotep397
- Start date
- Sort by reaction score
I've come to the conclusion that there just isn't enough demand for pro cards on the Mac Pro. It is frustrating that Nvidia and ATI offer dozens of different pro cards at different prices and capabilities that you can slap in any $600 quad core PC and none are availlable to use in the best workstation for the money. It's really infuriating. Same PC components inside, but you can't use any good video cards.
I see the Quadro 4800 as an experiment to test the waters so to speak.I've come to the conclusion that there just isn't enough demand for pro cards on the Mac Pro. It is frustrating that Nvidia and ATI offer dozens of different pro cards at different prices and capabilities that you can slap in any $600 quad core PC and none are availlable to use in the best workstation for the money. It's really infuriating. Same PC components inside, but you can't use any good video cards.
But there's a bit of a Catch-22. Software. If the software's not there, why would card makers really want to bother, and the software side is wanting the card makers to release, or at least make hard promises to do so, before writing the code to use it.
Maybe the Quadro will turn things around. Otherwise, you're stuck with Windows, and for that, you can use a PC card in a MP.
Ultimately, OS X user's interested will have to wait and see what shakes out.
I think there's more than sufficient demand, looking at the fact that the only people using DCC apps on Linux were studios back during the period I described...that's just not a lot of people comparing it to the entire market. I think it's more an issue of Apple's implementation being a bit of a moving target and not precisely what these companies want, so they're unwilling to bend or work outside of their existing development framework. Unfortunately, a lot of the hardcore UNIX graphics programmers moved away from the platform years ago except for pretty specific uses. My hope is that Apple would actually put a real UI on Darwin since most hardcore coding people want a GUI, but they tend to want a minimally invasive GUI. If Apple were to do that and actually make Darwin a accessible variant to more people there would be a better possibility of seeing something like independent GNU drivers for every PC video card that's made.
I've no idea of the actual demand. There's a few users here that certainly are, but I'm wary of basing any demand off of such few people. That's not to say it's not large enough, I just don't know the actual figures.
But from what I understand, those that need such support needed it "at that time", and were forced to go with another platform. And as you mention, Apple doesn't play well with others. Not for software or hardware (forced incompatibilities/lack of support in this regard; EFI firmware implementation, HDMI, MDP, and BluRay certainly come to mind).
But if the card makers see adequate demand, they're more likely to begin negotiations with OS X developers (3rd party), as they'd have to make the cards and drivers. Apple won't deal with it all all IMO, though if available, they may choose to offer it as a CTO option.
If it happens, it will be done by 3rd party developers and hardware makers inspite of Apple, as I do think it's possible.
To me companies using consumer hardware to determine whether there's a great need for professional hardware seems silly on it's face. They just don't want to invest the money, which would be fine, but with Darwin's limited use profile there isn't a good way for a company to just jump in and independently write drivers. The only plus I'm hoping for is that Intel's Larrabee might be easily SLI'd between 2 to 4 units and trounce the performance of pro cards available today via OpenCL. If that happens I won't care about Nvidia or AMD's offerings any more.
You'd be amazed at just how similar the professional cards are to the consumer units. Take the same GPU, add additional VRAM, modifiy the firmware, and develop new drivers. That's about it.
In practice, it does typically mean different PCB's to hold the additional VRAM, and maybe a larger ROM (or muliple ROM's) to contain the larger firmware. But the GPU, is the same if possible, due to the development and manufacturing costs. They can make different parts (i.e. change the address registers on it), but it's more cost effective to limit/allow features in the firmware.
It's just systems engineering to create the different models.
More work is placed on the software developers IMO, as they have to enable these features in their products, assuming they want to support it. But I'm under the impression, it can actually reduce their workload in the end (i.e. use some code to access the card's features, instead of trying to program their apps to do a specific function for a card that doesn't already support it (i.e like RISC vs. CISC instructions in days past). The compilers had to be developed to do decode complex instructions to a series of simple instructions in RISC based systems to perform the desired task.
Seems parallel to me. More work on the front end, but the payoff is worth it, and saves time in the long run. Ultimately, the users would certainly benefit, and it could give a notable edge. Especially in MP's, as Crossfire and SLI aren't supported to assist in brute force calculations.