Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Graphics cards and the New MacPro

MacVidCards said:
You traced 164 pins and followed the circuits? Through a multilayer PCB ? I'm sorry, that is total BS.

I have repaired damage to MANY GPUs in last 7 years, the traces are barely visible, then vanish from top layer to connect through an intermediate layer and show up 3 inches away from start at a component on backside. And it is in fact the backside that you have pulled this claim from.

Without multilayer schematics and hours tracing circuits there is 100% NO WAY to make any such claim .

I call "shenanigans"

In past Apple changed just a couple pins on AGP connector for use with ADC display and people have had to tape pins 3 & 8 on cards for G4s ever since.

I suggest you recant, or admit it's option 1 or 2.

Or explain how your X-ray vision allows you to see the inner PCB layers from jpegs.
Click to expand...

To be clear, I only looked at the general direction of where large groups of pins went and to what circuits they connected to. For example the first few pins go to a series of capacitors, which is consistent given that those are power pins, and the data pins travel to the processing core.

I didn't cross my mind that Apple would mess the PCIe 3.0 standard at all, but I will admit that we can't be 100% certain that everything is kosher just by seeing the back of the card, or even seeing the whole card if Apple messed with the pins before.

Does anyone know if Apple has ever messed with the PCIe standard?
 
2)
spaz8 said:
Ya I will be happy if we see an Nvidia option in the nMP. I worry it will just be AMD, but perhaps that was just marketing hype that OSX will finally have Firepro card support. OpenCL might be the future, but CUDA is today.
Click to expand...
My OpenCL code actually runs 9x faster on nV 570 than on AMD 5870! Have not tried it on GNC architecture. Curious about firepro performance; may have to rent one for a few days to see.

In other words, it's not necessarily either-or. If you need CUDA, then yes you do need nV. If you need OpenCL, you can use either, and either one may be (perhaps much) better for your task.

1)
Radiating said:
... ever messed with the PCIe standard?
Click to expand...
I doubt that they'd do that. Rearrange them perhaps, and add pins for their own purposes, but the PCIe signals, levels etc. should all be the same.

[I was going to reply this to one of your earlier posts on this thread:]
Yes of course someone will come up with a cable if TB bandwidth turns out to be a problem; hadn't thought of that! If the connector is proprietary they can still take a mold and reverse-engineer a mating connector, and figure out which pins are PCIe-relevant. It will cost more than a dollar, but still reasonable for those who really need it even if it means drilling a hole in the back i/o panel.
 
Radiating said:
To be clear, I only looked at the general direction of where large groups of pins went and to what circuits they connected to. For example the first few pins go to a series of capacitors, which is consistent given that those are power pins, and the data pins travel to the processing core.

I didn't cross my mind that Apple would mess the PCIe 3.0 standard at all, but I will admit that we can't be 100% certain that everything is kosher just by seeing the back of the card, or even seeing the whole card if Apple messed with the pins before.

Does anyone know if Apple has ever messed with the PCIe standard?
Click to expand...

mac_pro_2013_storage-580x460.png


434px-Mac_Pro_2013_inside.png


This connection between the new Mac Pro GPUs and the bottom circuit board does not look in any way like a standard PCIe slot to me (the black thing is mounted to the board - it's not like the board has a card edge inserted into a slot). Which is really not that surprising since Apple's design requires additional connections for the display port output and PEG power connections which a PCIe card slot does not facilitate. (And the GPU is piggy-backing the SSD). They most likely opted for a custom integrated connector to handle all of this.

BTW, PCIe bus interconnects can be implemented without using standard PCI Express card slots. This is done in Mac Books, iMacs, the Mac Mini, and millions of other laptops and computers that lack expansion slots.

Also, Apple's not the first vendor to consider offering dual high-end GPUs integrated into a main-board with TB... look up the ASUS Zeus motherboard.
 
Last edited:
VirtualRain said:
Yeah, I know... That's why I was careful to say "consider offering" in this context. It's interesting though and could support an argument both for and against the Mac Pros design.
Click to expand...

It turned some heads, but ultimately I think the anemic GPU choice and lack of upgradability turned everybody off (which is why it disappeared). Not really worth it, especially in exchange for 2 (count them, two--ah. ah. ah. ) Thunderbolt channels. I think the market is going to remain in upgradeable PCIe, at least for a while.

Gaming is a big driver:
http://www.ign.com/articles/2013/09/04/why-pc-gaming-has-exploded
http://jonpeddie.com/press-releases...g-hardware-sales-shrug-off-pc-market-decline/
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back