Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
I see no reason a citation is needed. Every device that can be daisy-chained will come with a HUB onboard, just as Firewire devices do now. A daisy-chain TB device will do the same. In fact, all the manufacturers have stated they will offer pass-through TB ports (2 TB ports). HUBs are implied to daisy-chaining. Powered HUBs, like all Firewire and USB HUBs before, will quickly join the TB party.

Hum... Hubs are not daisy chains. Hubs are part of a star topology, daisy chaining is a bus topology. Having 2 TB ports on a device is not having a hub.

So yes, Citation is needed to know if TB hubs can be produced to connect devices in a star topology instead of a bus topology. I have not seen it. Maybe it's out there, who knows, and that's why I asked for a citation before saying it's a fact that TB hubs and will exist.
 
Enclosure for what? If it's just a simple HDD then USB 3.0 would suffice (no benefits from Thunderbolt - just higher price). If it's SSD in RAID configuration (the only device that would benefit from Thunderbolt compared to USB 3.0) then it's price would be high anyways. But who will need those SSD/RAID monsters. Professional FCP users? Well, they all are switching to AVID anyways (we all know why).

Nobody that I've heard of has switched anything. FCP7 works fine until FCPX can fill the void.

Macs don't have USB 3.0 if you didn't notice.
 
The problem is not Thunderbolt. The problem is that Apple does not do USB 3.0. You may dream as much as you want about your Mac-size wealth. In reality with Thunderbolt and without USB 3.0 it's the Mac users who deserve a pity from PC side.

I build PC a Few motherboard now have USB 3, two ports only, no one cares. Just like TB USB 3 is new and for most users no one cares. Non of my customers care at all about it. They are more interested in SSD which makes a difference.

USB 3 a lot faster than USB 2 but let me know what right now and for say the next 24 months you would do with it?

Drop the Hard Drive off the list and its not that big a deal for anyone that is not a tech head. I am one of those but not my customers.

But I can give some places where they could be worth it but people can still live without.

1. High end professional cameras still no one is jumping over hoops for this
2. Professional Video cameras (could be important place for TB)
3. Very high end system for professional movie making
4. Servers (but not a must for now)

Places where no one except again tech head care.

1. Cheap computers
2. low level home use backup hardware and small business
3. Tablets
4. Gaming platforms
5. High end sound systems (Airplay I think will be the future here)

This thread is about AirPlay and Thunderbolt limited to High End devices.

Which makes sense, just like SSD for the most part is still on high end system though it has been coming down enough that the 128GB most people can afford, but take that to 200+ or even 500+ and how much of the consumer population will even touch it?

I will bet that 1 in 100 people will get any benefit out of TB, and maybe just maybe 10 out 100 for USB 3. That having TB and USB 3 everywhere would make older tech go away is just not possible, human psychology tells us people need to die off for real change. As I said before go online and check out how many bran new high end motherboards have still PS/2 port.

But if I am wrong I am open to been told so with explanation. :D
 
I build PC a Few motherboard now have USB 3, two ports only, no one cares. Just like TB USB 3 is new and for most users no one cares. Non of my customers care at all about it. They are more interested in SSD which makes a difference.

USB 3 a lot faster than USB 2 but let me know what right now and for say the next 24 months you would do with it?

Drop the Hard Drive off the list and its not that big a deal for anyone that is not a tech head. I am one of those but not my customers.

But I can give some places where they could be worth it but people can still live without.

1. High end professional cameras still no one is jumping over hoops for this
2. Professional Video cameras (could be important place for TB)
3. Very high end system for professional movie making
4. Servers (but not a must for now)

Places where no one except again tech head care.

1. Cheap computers
2. low level home use backup hardware and small business
3. Tablets
4. Gaming platforms
5. High end sound systems (Airplay I think will be the future here)

This thread is about AirPlay and Thunderbolt limited to High End devices.

Which makes sense, just like SSD for the most part is still on high end system though it has been coming down enough that the 128GB most people can afford, but take that to 200+ or even 500+ and how much of the consumer population will even touch it?

I will bet that 1 in 100 people will get any benefit out of TB, and maybe just maybe 10 out 100 for USB 3. That having TB and USB 3 everywhere would make older tech go away is just not possible, human psychology tells us people need to die off for real change. As I said before go online and check out how many bran new high end motherboards have still PS/2 port.

But if I am wrong I am open to been told so with explanation. :D

I can tell you what would do. In fact I am actually doing it. I have USB3.0 enabled external WD hard drive that I use for backup. Obviously it works much faster than anything else available on the market right now (for consumers).
 
You daisy chain Thunderbolt devices, you don't connect them to a central hub. Would you really like to connect a mouse to a keyboard to a hard drive to a printer to a tablet to a monitor? So if you want to take the printer out then you need to rewire your entire room?

Why would you not wifi a printer?

Why would you not bluetooth a mouse and keyboard?

2 printer wifi with static IP never had to worry about them, always work perfectly.

Mouse and key all bluetooth except the gaming machine which uses a USB 2 for that but that one machine of 5 at home. :D

Refrigerator is even WiFi. :) No I did not buy it because of it.
http://www.amazon.com/Samsung-RSG309AARS-Refrigerator-External-Dispenser/dp/B004XQHASW
 
As an Amazon Associate, MacRumors earns a commission from qualifying purchases made through links in this post.
I can tell you what would do. In fact I am actually doing it. I have USB3.0 enabled external WD hard drive that I use for backup. Obviously it works much faster than anything else available on the market right now (for consumers).

Thunderbolt > USB 3.0. You can't connect a fibre channel card via USB 3.0. Thunderbolt can already do this, even though the products haven't arrived to the market yet.
 
Citation is needed to know if TB hubs can be produced to connect devices in a star topology instead of a bus topology. I have not seen it. Maybe it's out there, who knows, and that's why I asked for a citation before saying it's a fact that TB hubs and will exist.

I see little reason why you couldn't have a thunderbolt connected breakout box with multiple ports on it. Could be USB or (e)SATA or ethernet or whatever. Would need some logic in the box to administer and aggregate the signals, that's all.

Thunderbolt is a PCIe interface, and multiport PCIe / USB adaptors are already widely available - they're called 'USB expansion cards for desktops'.

Personally I'd probably use mine for plugging in a really fast SSD drive. Already two fast SSD drives in RAID1 can theoretically overwhelm Thunderbolt :eek:

[but an enclosure that can process 1GB/sec of RAID 1 *might* be expensive]
 
I don't get it !

What is the hold up?

If an external drive enclosure now costs say $50 and a 1TB drive costs $100, and it costs another $100 to add Thunderbolt, why can't we see Thunderbolt drives for $250???

I would pay $250 for a fast Thunderbolt drive.

Am I missing something?
 
Thunderbolt > USB 3.0. You can't connect a fibre channel card via USB 3.0. Thunderbolt can already do this, even though the products haven't arrived to the market yet.

Exactly. Right now you can not connect anything to Thunderbolt. And now we are told that even if there will be any products, they will be more expensive than the ones for USB 3.0 (with no speed benefits compared to USB 3.0 for most typical applications).
 
2. Professional Video cameras (could be important place for TB)

TBolt is not a network or peripheral protocol. It is a PCIe expansion port that routes a small number of PCIe lanes out of the system chassis.

There are *no* TBolt devices available or proposed. Every proposed device has a TBolt to PCIe bridge, and a PCIe controller to connect to the actual device protocol.

The LaCie disk has
TBolt -> PCIe -> SATA controller -> SATA drives​
inside of it. Is it any wonder that they're afraid to announce the price? (Note that you also have to include the second TBolt port for daisy-chaining.)

I really don't see camera manufacturers adding all that hardware inside a camera, plus two mid-sized port connectors - when USB 3.0 is probably far faster than the flash drive embedded in the camera. Lots of pain ($$) for no gain.

And, USB 3.0 is compatible with far, far, far more systems that TBolt ever will be.
 
I see little reason why you couldn't have a thunderbolt connected breakout box with multiple ports on it. Could be USB or (e)SATA or ethernet or whatever. Would need some logic in the box to administer and aggregate the signals, that's all.

Because a BUS topology technology does not work the same as a Star topology technology.

Again, if it is possible, please provide some documentation showing it is. Otherwise, qualify any statements of "TB hubs are coming" by stating that this is your opinion based on your own speculation. Don't state it as fact.


I don't get it !

What is the hold up?

If an external drive enclosure now costs say $50 and a 1TB drive costs $100, and it costs another $100 to add Thunderbolt, why can't we see Thunderbolt drives for $250???

I would pay $250 for a fast Thunderbolt drive.

Am I missing something?

Not big enough of an installed base to justify the design of the thing. The volume would be pityful right now for any vendors, so low cost, low-end solutions aren't worth designing and bringing to market. Chicken and egg problem.

USB3 gets around that with backwards compatibility.
 
This phenomenon is just so much like Apple. You can only get Mac computer at approximately $1000 for the lowest end

Or a MacPro for minimum $2500. If u think it all over again, it's not that Mac being overpriced, they just don't make the cheap version of their Mac.

And so does thunderbolt, being $100 more than the competition without thunderbolt, but offer great improvements??! It's not exactly overpriced, see? But available only at $1000++ device while a simple $60 1Tb external hdd could implement thunderbolt too, if they just allow it :p
 
I can tell you what would do. In fact I am actually doing it. I have USB3.0 enabled external WD hard drive that I use for backup. Obviously it works much faster than anything else available on the market right now (for consumers).

Yes I understand that, but we are talking about consumers not a limited demographics. For most consumers their backup HD does not even exist, so putting TB or USB 3 on cheap equipment just to say it has it is is not a priority.

Give it another 12 months and I see way more things coming out.

Personally I want TB flash drive :rolleyes: but that just me or a select limited customer. So having TB and USB 3 limited to high end for the next year or two probably not going to impact anyone.
 
Personally I want TB flash drive :rolleyes:

That would be a humongous flash drive, what with the stand-along TB controller, the 2nd port for a daisy chain and all the added chips to enable a PCIe bus and a device controller on top of the flash memory itself. :eek:

I'd rather have a USB 3 one myself for half the price, with twice the storage at half the size... ;)
 
Price per MB/s

Source Lacie website:

Thunderbolt 4TB: ~$1,400 = .35/MB up to (10,000 MB/s)

eSATA 2TB: ~ $250 = .13/MB up to (480 MB/s)

USB 3.0 1 TB: ~ $210 = .21/MB up to (110 MB/s)

Clearly the choice is what best fits your needs. Thunderbolt technology is the future.

USB 2 and 3 will continue be a slow low, end and cost solution for the average throughput and user.

Why have a short sighted view point. Demands of software and the average user is only going to increase as it always has.

I will take the speed Please.
 
Last edited:
Screw that, USB 2.0 is fine for me, don't care if it takes a bit longer to copy a file. This will either be everywhere at no extra cost in 5 years, or it will be unheard of just like like Mini DisplayPort.
One file isn't a problem. It's when you are copying gigabytes of raw video footage where the speed counts. Think of copying data to an external hard drive faster than your internal hard drive.... this is also designed to allow one-cable connection from a laptop to a workstation (drives, hub, monitor, sound, recording equipment).
 
One file isn't a problem. It's when you are copying gigabytes of raw video footage where the speed counts. Think of copying data to an external hard drive faster than your internal hard drive....

Wait, faster to an external HD than an internal one ? :rolleyes:

You do understand there's 0 difference between a Thunderbolt connected external drive and a SATA 3.0 connected internal drive right ? There's no magical pixie dust.
 
I see little reason why you couldn't have a thunderbolt connected breakout box with multiple ports on it. Could be USB or (e)SATA or ethernet or whatever. Would need some logic in the box to administer and aggregate the signals, that's all.

Thunderbolt is a PCIe interface, and multiport PCIe / USB adaptors are already widely available - they're called 'USB expansion cards for desktops'.

Personally I'd probably use mine for plugging in a really fast SSD drive. Already two fast SSD drives in RAID1 can theoretically overwhelm Thunderbolt :eek:

[but an enclosure that can process 1GB/sec of RAID 1 *might* be expensive]

Do you mean "RAID-0" instead of "RAID-1". A RAID-1 drive will be slower than a single SSD at writing. If the controller is smart and you have lots of concurrent reads, RAID-1 can be faster than a single drive. RAID-0, however, is often close to twice as fast as a single drive for large reads and writes, and often significantly faster for small reads and writes.


Because a BUS topology technology does not work the same as a Star topology technology.

I may be confused about the total thread context, but the post that you replied to doesn't involve star vs. bus technologies.

A TBolt device that had multiple outputs (eSATA ports, USB 3.0 ports, GbE ports, 1394 ports) could be implemented with a simple PCIe->PCIe bridge (in essence a PCIe hub), with multiple PCIe controllers inside the "docking station". To the TBolt controller, this would look like a single PCIe device - but to the host it would look like four separate PCIe devices. (There's an assumption here that a TBolt device could have a PCIe->PCIe bridge in it. We have no idea if that is permitted.)

I agree completely with you, however, that we've seen nothing to suggest that TBolt will support tree topologies. (We've seen darn little technical info on TBolt, actually.) Can you put a 4-way PCIe-to-PCIe bridge on a TBolt controller, and run four TBolt chains from that device? No idea, I want to see the citations too.
 
I may be confused about the total thread context, but the post that you replied to doesn't involve star vs. bus technologies.

The sub thread started about TB hubs, ie, 1 TB output to a computer, multiple TB inputs from devices, not TB -> SATA/GbE/1394 ports. Basically, allowing you to connect devices without first unplugging the rest of the chain or the external monitor that is run off DP.

As you say, I've seen and heard nothing of the sort as a possibility for TB. It's all been "daisy chaining" and "2 ports per device, 1 input, 1 output".
 
Hum... Hubs are not daisy chains. Hubs are part of a star topology, daisy chaining is a bus topology. Having 2 TB ports on a device is not having a hub.

So yes, Citation is needed to know if TB hubs can be produced to connect devices in a star topology instead of a bus topology. I have not seen it. Maybe it's out there, who knows, and that's why I asked for a citation before saying it's a fact that TB hubs and will exist.

Firewire and USB are not sequential bus protocols. They poll their bus just as a Star topology polls its network. And two pass-thru ports, wether FW, USB, or TB are HUBs. They are basically dumb HUBs. The controller does the polling not the ports. In my Firewire chain of HHDs I could access the last in line even with the the one before completely off.
 
please, "TB" has meant "TeraByte" for years

The sub thread started about TB hubs, ie, 1 TB output to a computer, multiple TB inputs from devices, not TB -> SATA/GbE/1394 ports. Basically, allowing you to connect devices without first unplugging the rest of the chain or the external monitor that is run off DP.

As you say, I've seen and heard nothing of the sort as a possibility for TB. It's all been "daisy chaining" and "2 ports per device, 1 input, 1 output".

Can we please agree to say "TBolt" for "Thunderbolt" - and keep "TB" to mean TeraByte as it has for years?

When one is talking about TeraByte Thunderbolt disks - it gets confusing!
 
Can we please agree to say "TBolt" for "Thunderbolt" - and keep "TB" to mean TeraByte as it has for years?

When one is talking about TeraByte Thunderbolt disks - it gets confusing!

What's wrong with a 2 TB TB disk ? :D

Firewire and USB are not sequential bus protocols.

Good, we're talking Thunderbolt here though. Got any links to documentation about the possibility for Thunderbolt ?
 
Firewire and USB are not sequential bus protocols. They poll their bus just as a Star topology polls its network. And two pass-thru ports, wether FW, USB, or TB are HUBs. They are basically dumb HUBs. The controller does the polling not the ports. In my Firewire chain of HHDs I could access the last in line even with the the one before completely off.

But not if the one before is unplugged!

:p

Seriously, though, "bus power" was included in the 1394 spec to make this possible. It isn't there to power the device itself, it's there so that a 1394 controller has enough power available to pass through signals if the main device is off.

And your other statements may not be wrong, but until we see some official announcements of TBolt hubs we should not assume that they are on the way.


What's wrong with a 2 TB TB disk ? :D

Wouldn't it be a 2TB² disk?
 
I still fail to understand why Apple didn't include USB 3.0 on the 2011 MBP.

As a regular consumer, I'll never see the performance increase from 5Gb/s to 10Gb/s. However, I sure do see not having any reasonably priced peripherals.


For all practical purposes, Apple might as well not be including thunderbolt in their new computers. Thunderbolt is utterly worthless today, and at least for immediate future. Maybe in 2012 or 2013 thunderbolt may be worthwhile. It sure as heck isn't worth anything today though...


What's kind of frustrating for this whole scenario is that classically Apple is really good at making a good consumer experience, and cares less about technical specifications. You can find a Dell laptop that is technically faster than a MBP but just isn't as fun to use. When Apple chose Thunderbolt over USB 3.0, it acted like a PC.
 
Enclosure for what? If it's just a simple HDD then USB 3.0 would suffice (no benefits from Thunderbolt - just higher price). If it's SSD in RAID configuration (the only device that would benefit from Thunderbolt compared to USB 3.0) then it's price would be high anyways. But who will need those SSD/RAID monsters. Professional FCP users? Well, they all are switching to AVID anyways (we all know why).

For PCI and PCIe expansion chassis perhaps, running a Pro Tools HD rig on a laptop? Why not. From intel product description:

"We are very excited by the capabilities of Thunderbolt technology. To have two 10Gbps, bi-directional, multi-protocol channels in a single cable is a great step forward for high performance audio and video solutions."
– Max Gutnik, Sr. Director, Product Management, Avid Technology


Or, how about connecting computers as nodes in a cluster? Yeah, access to the PCIe bus is pretty cool. Or how about daisy chaining multiple high resolution monitors with one cable?

What's funny is that re FCP X, comments arise that Apple don't care about pro users, it's all consumers nonsense. Then at the same time they adopt a truly professional format that has great potential for high bandwidth data transfers at low latency. It's a great successor to the ExpressCard, which pro users cried about when it disapeard from the 15" macbook pro.

One interesting point about usb3 is that intel themselves have not yet made a chipset for it, it's all 3rd party implementations so far.
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.