Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Lets not speculate, there is no documentation that it COULDN'T be possible, and there certainly is no reason other than your own pessimism to suggest that it wouldn't work.

So you're proposing a religious argument instead of a scientific one. When you lack evidence, you assume the option is non-existent. To do otherwise is to hold faith. I only argue facts, not faith since with Faith and speculation, I can make it so Thunderbolt ports have money come out of them. Yes, big bills, 1000$ per day type stuff, freshly printed and pulled out the port. After all, there is no documentation that it COULDN'T be possible.

Technology is about reason and logic, not faith. Drop the blind faith if you want to have a worthwhile discussion.

Actually, we do have some evidence that it is possible.
Things we know:
1) From a protocol stand-point, Thunderbolt is essentially PCI-E over an external cable.
2) The 27" iMac has *two* Thunderbolt ports with full capabilities on each

Both of which do not indicate that Thunderbolt supports anything but BUS type configurations. The 2 ports are simply seperate BUSes endpoints and from a protocol stand-point, PCI-E is a BUS type architecture, not a star type architecture.

Again, let's talk actual facts, not idle faith based speculation (it's based on PCI-E so you can have a hub!). Provide citations if you're going to claim something that is not presented as such by Intel and current reviewers.
 
So you're proposing a religious argument instead of a scientific one. When you lack evidence, you assume the option is non-existent. To do otherwise is to hold faith. I only argue facts, not faith since with Faith and speculation, I can make it so Thunderbolt ports have money come out of them. Yes, big bills, 1000$ per day type stuff, freshly printed and pulled out the port. After all, there is no documentation that it COULDN'T be possible.

Technology is about reason and logic, not faith. Drop the blind faith if you want to have a worthwhile discussion.
lmao, saying a hub is possible is NOWHERE in the realm of saying the port will print $1000 bills.... you are really reaching for straws with that one.

TB is supposed to not care about protocols, it is adaptable to everything, so why would it not be adaptable to a hub? If you could provide ONE reason other than Intel hasn't said it can, then I'll play along. Until then you are just pulling **** out of your ass to further degrade TB. If you don't want to use it, guess what? YOU DON'T HAVE TO.... Apple didn't take anything away from you by implementing the port. You can still use it as ONLY a display port for your monitor and continue using all of your other current drives and accessories. Apple isn't forcing you to switch all of your stuff to TB technology, they aren't forcing you to daisy chain drives, they aren't forcing you to buy adapters, it's still the same freaking port that's been on macs for years now, only IF you CHOOSE, in the future you will be able to plug other things into that same port. Again if you don't like it, that's fine, don't use it, but you act as if apple has somehow screwed you by putting TB in the macs... guess i just don't see this logic.
 
lmao, saying a hub is possible is NOWHERE in the realm of saying the port will print $1000 bills.... you are really reaching for straws with that one.

It's based on blind faith and speculation. Not any kind of fact. I'm not reaching, I'm hyperboling so you understand how fictous your hub scenario is, until it is proven to be fact.

USB is also protocol agnostic. I can run Audio, network, graphics, storage over it. All these employ their own protocols. USB just encapsulates it in a lower USB layer. This is not inherent to TB. TB doesn't understand Ethernet or USB or PCM encoded audio. It will require transceivers on both end that do. TB is a transport bus. It moves data fast.
 
I've never used FireWire/IEEE 1394/iLink. Do they play well together?

Yes, except that the iLink device won't be powered by the bus. Offhand, I don't recall seeing any iLink using device having more than one port, so they end up being the terminal device in the chain.
 
Btw, sony did exactly the same there as apple with (m)dp.
First they made a proprietary solution, then standardized it to 1394a in 2000.

With the minor difference that miniDP is functionally equivalent to DP, while 4-pin iLink is a subset of standard 6-pin, and doesn't supply bus power to its device.

Fairly major functional difference in intent.
 
It's based on blind faith and speculation. Not any kind of fact. I'm not reaching, I'm hyperboling so you understand how fictous your hub scenario is, until it is proven to be fact.

USB is also protocol agnostic. I can run Audio, network, graphics, storage over it. All these employ their own protocols. USB just encapsulates it in a lower USB layer. This is not inherent to TB. TB doesn't understand Ethernet or USB or PCM encoded audio. It will require transceivers on both end that do. TB is a transport bus. It moves data fast.
So if I can add all of those transceivers to a cable, why wouldn't I be able to build a box with the same transceivers that allows me to plug in whichever one I want (usb, firewire, ethernet, tb, etc.)?

You are really just being a pessamist trying to find every little thing you can about TB to complain about. You claim it's going to screw your display by having video/data on the same cable, but there is a seperate 10Gbps data channel for PCI-E and DP so that won't be an issue.

Your issue with the monitor being last is probably relevant for, well, pretty much nobody.... How many people need or can even afford to have 3 TB drives between their computer and monitor? And if you do have 3 TB drives hooked up, how many situations do you need to pull out a drive in the middle? If you are going to have to pull out a drive you should know it ahead of time, it's not like all of a sudden you are just going to need the second drive in the chain for some reason. Plan ahead and put it as the last one on the chain, then even if your monitor is last because you don't want to purchase a new one, you have to unplug 1 monitor to dismount your drive then plug it back in... hardly an inconvenience if the situation presents itself. You can keep harping on well what about the 1 time a year i might need to do this and i will be inconvenienced for 30 seconds or you can look at all the positive aspects of what thunderbolt brings to the table. I really want to know what situation you are going to have so many problems because your 4 tb SSD RAID arrays are between your first and second external monitor and you suddenly need to pull the raid array out.

Your attitude is honestly quite annoying, you just read all of the articles and look for what you can bitch about apple doing today? Again, nobody forces you to use tb if it's that much of a pain in the ass for you, use the same technology you are using today... nobody took that ability away from you.
 
Okay, so I have an honest question. There is a lot of talk about the bandwidth limits and the fact that ThB cuts the throughput bandwidth of the mDP in half (10 Gbps vs. 20 Gbps). But, why does one need so much throughput, except in really extreme cases? By my calculation, the Apple 27" Cinema display has about 3.5 mega pixels. At that count, it would take a connection speed of about 2 Gbps to drive the display at 16 bit color depth, at a frame rate of about 30 fps. And that's assuming that you're updating every pixel with every frame.
 
Okay, so I have an honest question. There is a lot of talk about the bandwidth limits and the fact that ThB cuts the throughput bandwidth of the mDP in half (10 Gbps vs. 20 Gbps). But, why does one need so much throughput, except in really extreme cases? By my calculation, the Apple 27" Cinema display has about 3.5 mega pixels. At that count, it would take a connection speed of about 2 Gbps to drive the display at 16 bit color depth, at a frame rate of about 30 fps. And that's assuming that you're updating every pixel with every frame.

You are still missing out on the multiple video streams that DP 1.2 would allow. Right now it is 1 monitor per port. With DP 1.2 it is possible to run 4 1080p monitors off 1 port. Bandwidth just helps drive higher resolutions with a higher refresh rate.
 
You are still missing out on the multiple video streams that DP 1.2 would allow. Right now it is 1 monitor per port. With DP 1.2 it is possible to run 4 1080p monitors off 1 port. Bandwidth just helps drive higher resolutions with a higher refresh rate.
yeah but there are adapters that will allow you to run dual displays off of one port.
 
So if I can add all of those transceivers to a cable, why wouldn't I be able to build a box with the same transceivers that allows me to plug in whichever one I want (usb, firewire, ethernet, tb, etc.)?

Yeah, why can't you just plug 10Base2 in a star topology, it just requires a box with the proper transceivers right ? Because that's not how BUS technologies work plain and simple.

Until Intel documents some method of enabling star/tree topologies with TB or it is shown to be possible, then it is not possible. End of story.

yeah but there are adapters that will allow you to run dual displays off of one port.

But you're still losing out on half the bandwidth unless you use both TB channels independantly, then you're stuck that all your traffic is going the same way, out from the computer to the display. Too bad you're also trying to read data from a HDD back into the computer...

We're going in circles. Look, TB and DP don't mix well. It's a big compromise.
 
Yeah, why can't you just plug 10Base2 in a star topology, it just requires a box with the proper transceivers right ? Because that's not how BUS technologies work plain and simple.

Until Intel documents some method of enabling star/tree topologies with TB or it is shown to be possible, then it is not possible. End of story.
alright, maybe it is, maybe it isn't possible, not the end of the world if it's not still don't know why you're complaining but I don't care to argue this one with you anymore because there is really no evidence on either end.

But you're still losing out on half the bandwidth unless you use both TB channels independantly, then you're stuck that all your traffic is going the same way, out from the computer to the display. Too bad you're also trying to read data from a HDD back into the computer...

We're going in circles. Look, TB and DP don't mix well. It's a big compromise.
there are separate data channels for PCI-E and DP in thunderbolt so you are splitting your 10Gbps DP channel in half to run 2 displays and your data is still working just fine at 10Gbps on PCI-E data channel. I think that 5Gbps would be fine for each of two external displays, no?
 
there are separate data channels for PCI-E and DP in thunderbolt so you are splitting your 10Gbps DP channel in half to run 2 displays and your data is still working just fine at 10Gbps on PCI-E data channel. I think that 5Gbps would be fine for each of two external displays, no?

Wait, what ? Now we're down from 21 Gbps to 5 Gbps ? Wow. It's even worse than I thought.
 
Coupled with the fact that DP 1.1 only allows display mirroring (from a single port)...

I already went over that point : Apple does not even support the latest DP spec...

Seriously. I think Sony is on to something. Keep the displays seperate from all the other data based junk.
 
...

So, anyways....
since there are no thunderbolt peripherals it is a completely useless port at this point. When peripherals do come out, the connector they use will obviously win. Unitl that point in time it is all a pipe dream. This would not be the first time apple lost a IO competition (firewire/usb) so we might all just look forward to usb 3 adapters so we can actually use the faster port.
 
With the minor difference that miniDP is functionally equivalent to DP, while 4-pin iLink is a subset of standard 6-pin, and doesn't supply bus power to its device.

Fairly major functional difference in intent.
Nope,
http://en.wikipedia.org/wiki/Firewire#Improvements_.28IEEE_1394a-2000.29
:"The 4-circuit connector is fully data-compatible with 6-circuit alpha interfaces but lacks power connectors."
Considering the battey tech at the time, it is no brainer to have no-power connection for mobile devices.
Surprisingly the original spec was maybe intended only for desktop devices.
 
Wait, what ? Now we're down from 21 Gbps to 5 Gbps ? Wow. It's even worse than I thought.
what are you talking about? There are 2 bidirectional lanes on each channel (PCI-E and DP) each can carry 10Gbps so you have a total of 40Gbps of data transfer.

Since each channel is bidirectinal you have 20Gbps up AND 20Gbps down. However a single device can't "combine" them so each device has a max of 10Gbps. Therefore:

2 displays, each could use one lane of the DP channel at 10Gbps.4 displays means each would have to use only 5Gbps because I don't think the DP can use the PCI-E channel and vice versa.

Now I know you're going to come back and say "yeah so it's half of the 20Gbps of DP" well each of 2 displays can only use "half" but before you ask the question, what displays are you running that EACH need more than 10Gbps?
 
Okay, so I have an honest question. There is a lot of talk about the bandwidth limits and the fact that ThB cuts the throughput bandwidth of the mDP in half (10 Gbps vs. 20 Gbps). But, why does one need so much throughput, except in really extreme cases? By my calculation, the Apple 27" Cinema display has about 3.5 mega pixels. At that count, it would take a connection speed of about 2 Gbps to drive the display at 16 bit color depth, at a frame rate of about 30 fps. And that's assuming that you're updating every pixel with every frame.
Display interfaces are not so intelligent that you have to refresh only those pixels that change from GPU. It's every pixel every time.
And there is 3 colors in 1 pixel, so minimum is 24 bits per pixel. Win7 (most advanced OS...?) already supports 10 bits per color (30 bits per pixel) like many new high-end monitors.
Both dp & hdmi also supports 16 bit colors (48 bits per pixel).

So lets go forward 3 generations of display tech advancements.
Lets say that HDR becomes so mainstream that we start to use 16 bit colors.
And lets assume 3D is very common.
And ACD's new version is once again little bigger from 2560x1440 to real "4k" aka 4096x2160.
For stereographic 60Hz 4k HDR picture you need:
4096 * 2160 * 16 * 3 * 60 * 2= 50,960,793,600 bit/s = 51 Gbit/s!

This is of course only for those geeks that try to convince you that thay make hollywood blockbusters with their mbp's because they know it tech, but gives you the hint what's coming.

Even a normal guy like 3D artist, architect, ind. designer or doctor might like to have double 3D 2k screens with 12 bit color depth. That can be achieved with dp1.2.
But for 27" ACD resolution that would need dp1.3.

Of course if you keep re-buying your whole it department every year, this matters you nothing.
 
Okay, so I have an honest question. There is a lot of talk about the bandwidth limits and the fact that ThB cuts the throughput bandwidth of the mDP in half (10 Gbps vs. 20 Gbps). But, why does one need so much throughput, except in really extreme cases? By my calculation, the Apple 27" Cinema display has about 3.5 mega pixels. At that count, it would take a connection speed of about 2 Gbps to drive the display at 16 bit color depth, at a frame rate of about 30 fps. And that's assuming that you're updating every pixel with every frame.

DP 1.2 is not extreme.

DP 1.1a cannot drive 2x 30" 2D displays from 1 port.
DP 1.1a cannot drive 1x 30" 3D monitor.

DP 1.2 can.
 
what are you talking about? There are 2 bidirectional lanes on each channel (PCI-E and DP) each can carry 10Gbps so you have a total of 40Gbps of data transfer.

Since each channel is bidirectinal you have 20Gbps up AND 20Gbps down. However a single device can't "combine" them so each device has a max of 10Gbps. Therefore:

2 displays, each could use one lane of the DP channel at 10Gbps.4 displays means each would have to use only 5Gbps because I don't think the DP can use the PCI-E channel and vice versa.

Now I know you're going to come back and say "yeah so it's half of the 20Gbps of DP" well each of 2 displays can only use "half" but before you ask the question, what displays are you running that EACH need more than 10Gbps?
um, the video portion of DisplayPort is not bi-directional.
 
what are you talking about? There are 2 bidirectional lanes on each channel (PCI-E and DP) each can carry 10Gbps so you have a total of 40Gbps of data transfer.

No, each Channel can carry 10 Gbps. There's 2 Channels in TB. What are you talking about ?

TB is 20 Gbps bi-directional, but guess what, DP is uni-directional. Monitors don't push back that much data to a computer.

You've been confused this whole thread.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.