It's capable, but it cannot be used in a star topology.
Can you even start to imagine how much a Thunderbolt hub would cost ?
299$ for a Belkin generic POS port extender with cheapo ports... now imagine a 4 Thunderbolt port affair...
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Thunderbolt Adoption Reportedly Slowed by Intel's Licensing and Certification, But Improvements Coming
- Thread starter MacRumors
- Start date
- Sort by reaction score
Can you even start to imagine how much a Thunderbolt hub would cost ?
299$ for a Belkin generic POS port extender with cheapo ports... now imagine a 4 Thunderbolt port affair...I forgot you had mentioned the fact that it worked fine when you weren't running Mac OS. Definitely sounds like Mac OS is scaling or not reading the full EDID. That is annoying. I assume you've futzed with the underscan settings in the display control panel?
Can you even start to imagine how much a Thunderbolt hub would cost ?
You've complained about daisy chaining, but then went on to essentially explain why it exists as a solution in the first place. Ethernet ditched 10BASE-2 as soon as 10BASE-T hubs were cheap enough that people could afford them. People will often balance the cash outlay they are capable of vs. the inconvenience level they can tolerate.
You're not going to get very far with a 10 Gbit/s hub (collisions much?). A Thunderbolt switch is possible but would cost about $100 per port or more. Very few people would opt to spend $550 to $850 at this point just to avoid the inconveniences of daisy chaining.
And the Belkin dock is not a port extender. It's a PCIe GbE NIC, PCIe FW800 host adapter, PCIe 4-port USB 3.0 HBA, and USB audio device in a single small enclosure that connects to the host PC and other accessories/displays via a 2-port Thunderbolt controller. I have no idea if it will ever reach market or be any good once it gets there, but from a hardware standpoint, it has virtually nothing in common with port extenders/replicators or a Thunderbolt switch.
Yeah tried all sorts. Even resorted to trying someone else's screen, with no luck. I gave up in the end as I'm happy to use it over MiniDisplayport > DVI.
one cable to my computer, that means one to disconnect if I want to take my portable non of the other need to be futzed with..
Can you even start to imagine how much a Thunderbolt hub would cost ?
The whole point is that you do not need a hub, a hub would also saturate the bus since all devices added to the hub would share the bus going from the hub. Thunderbolt makes it easy to extend and add peripherals that needs the bandwidth on computers that lacks pcie card slots.
Exactly, it makes it easier to extend and add stuff without the need for extra devices.
There's only 1 cable to your computer in a star topology too... you know, the one going to the "breakout box". 1 Ethernet cable for a switched Ethernet network, 1 USB cable going to the USB hub... etc.. etc..
Sorry, still don't see the advantage of a Daisy chain in your post here...
----------
Buses are most notorious for collisions, but then again, that's a protocol design, not a topology problem really. 10Base2 was much more prone to collisions than hubs in star topology. So that's not the reason why "switches" would be required.
As for "switch ports being expensive", the daisy chain in Thunderbolt doesn't make it cheaper. For one, devices that can be chained require 2 ThunderBolt ports, which prevent the use of the cheaper controllers (Port Ridge 1 channel controller). So in essence, the hardware is already there, each device needing an expensive controller anyhow. So cost isn't really an issue.
No frankly, I don't know what Intel was thinking in even making Daisy chaining available, much less desirable.
Well, I was thinking of a portable hard drive situation where you friend/family member might come over and want a copy or want to share their copy of family photos or home videos or whatever under the sun. The point is that they would need to connect to your computer and then when they're done they're going to disconnect the drive and take it with them. You, meanwhile, have to fiddle with your monitor connections, etc. to accommodate them with daisy-chaining. A hub is SO much preferable. Now I'm not against a daisy-chaining OPTION, but so far I see no Thunderbolt hubs/switches for just more ports and for some odd reason I doubt I ever will (technical issues and/or price?)
I might not "need" one, but I sure as heck would WANT one for the aforementioned reasons (i.e. daisy-chaining is a ROYAL PITA).
As I've said in my previous post, there's nothing "easy" about daisy-chaining. Connect/disconnect devices fairly often with a monitor on the end and you'll quickly find it's a ROYAL PITA really fast. This is why all Macs should have at least TWO TB ports (i.e. one for monitors and one for data and if you need two for one or the other you'd have them, but at least you wouldn't likely interrupt your monitor chain as often as data and more importantly a come/go type hard drive or whatever could always just be the last device in a chain of home equipment (i.e. it's not a pain IF temporary devices are plugged in LAST). The problem is that monitors that do not provide their own pass-through (i.e. ALL monitors except Apple's with the TB hub on it) *MUST* be the last device and that's were it goes South very quickly.
Why would you want one if you don't need one? Daisy chaining is perfect for the intended purpose, you can extend without the need of extra gear, and it's simple and maintain full bandwidth between devices.
That's just your opinion, you would have your devices between the monitor and your computer in this case, also the alternative would be far more expensive for the intended purpose (although I agree that two ports would be ideal, which I believe some models have). What are you going to do, get yourself some infiniband instead? That'll be cheap.
Look, all current Macs now have usb3 so what is the problem, use usb3 if you like. Thunderbolt is not there to replace usb and the port functions as a display port as well if you do not use it. If you have a rack with some gear in, it's great to keep everything in the rack, it's been a main advantage of firewire as well for example, rack up a converter and add a hard drive to the rack. Look what companies like Avid is saying about Thunderbolt, it can be used for accelerators and dsp boxes with direct attached storage and so much more of that nature.
Last edited:
Or you could just put it on a thumbdrive and had it to them. or burn it to a CD so that can take it with them..I cary a 16GB thumb drive on my key chain exactly for this reason
I don't think I would ever hand someone one of my HDD friends/family or not not all info on the drive is for public consumption..
----------
Your not ever going to see the advantage and that fine too..
Oh, I dunno. Try reading my previous posts to find out, maybe?
NO, IT ISN'T. Try reading my previous posts. It's a PAIN IN THE BACK SIDE.
Gee, I thought that's what a discussion board was all about.
You can have fun unplugging and plugging cables all day long if you want. It's a load of crap, IMO. It's the same thing as running out of ports on a USB hub. You have to unplug something to plug something else in, except in this case you get to keep the device working. But when I have to do that with a USB hub, it means it's time to get another or a larger hub.
Yeah, I think I said that two posts ago. Unplug your monitor (or in the case of multiple hard drives, go look for one wherever it's stored on your desk and disconnect it to something else and insert your device in the middle...lots-o-fun(tm).
Hence the reason some of us are saying that daisy-chaining sucks. But to answer your question, IF there were a $50 4-port (or better yet 7-port) Thunderbolt hub, YES, it would be worth every penny (assuming I ever had enough devices to bother; right now all the TB equipment out there is so freaking expensive I'd NEVER buy it. Thus in reality, I need a good USB3.0 hub that works with Macs properly (a totally different thread, however). That's because I have TWELVE USB devices (and currently only 11-ports) and that's not counting temporary things like USB stick drives and meanwhile, the only thing plugged into my Thunderbolt port is my Mini-display port to HDMI adapter ($3) so I can use my second monitor with my Mini.
It's just an example, dude. I don't care what the devices are. The point is that IF you need to change them, it becomes a pain in the butt very quickly and hence my argument against daisy-chaining. I have USB hubs and banks of Ethernet switches in my house and IF I had to daisy-chain them instead, I think I would go insane.
Oh, I dunno. Try reading my previous posts to find out, maybe?
NO, IT ISN'T. Try reading my previous posts. It's a PAIN IN THE BACK SIDE.
Typing with caps lock is not going to convince me, I have personal experience with daisy chaining from scsi, firewire and midi and it's great for it's intended purpose. Perhaps the intention is not clear to you.
Seeing as it's intended purpose is to be a pain in the arse, I'd say you're correct.
I've got daisy-chaining experience dating back to the 1541 disk drive on the C64. In fact, we used to daisy-chain three or four at a time at users group meetings sometimes. But then we had to drag our drives there anyway so it was already inconvenient. Later on with the Amiga, we had to bring our own computes as well so a little daisy-chaining was no big thing (it daisy-chained floppies up to 4 also and my 3000 had SCSI which introduced even more of a PITA with device termination added to the mix).
These devices are the reason I have an opinion about daisy-chaining to begin with. I didn't like it then. I don't like it now. Yes, it's better than NO option. But a hub/switch with a boat load of ports beats it hands down any day of the week for convenience. You're also dependent on the device manufacturer for any given device to provide a (working) pass-through or the chain stops there. And that's precisely the issue that occurs with most monitors becuase Apple decided to combine mini-display port with Thunderbolt. It means every monitor out there (save Apple's TB one) will need to be the last device in the chain, making it a royal PITA again. You have a bit more flexibility when ALL your devices have pass-throughs since most temporary devices would just go to the end of the chain. All 1541 drives had a pass-through for example. But monitors (never designed to be in a TB chain) do NOT.
Seeing as it's intended purpose is to be a pain in the arse, I'd say you're correct.
These devices are the reason I have an opinion about daisy-chaining to begin with. I didn't like it then. I don't like it now. Yes, it's better than NO option. But a hub/switch with a boat load of ports beats it hands down any day of the week for convenience.
Picking a different solution that requires switches and hubs would be more expensive while maintaining the purpose (high bandwidth, low latency) of Thunderbolt. It's not a usb replacement, if you like usb and it's enough for your needs, use usb.
Some conveniences are worth a few extra bucks. It's so sad you cannot comprehend that simple fact. Why have a pull cord mower when you can have key-start? Why use a fan when you can have air-conditioning? Sorry, but I didn't buy a Mac to sweat my arse off, so to speak.
Fortunately, like I said, I only use the port for a 2nd monitor and given the utterly ridiculous prices of TB peripherals, it WILL stay that way. The price of a switch/hub for TB is irrelevant until a device comes along I need that can actually use that bandwidth. Otherwise, USB3 is more than adequate for mundane stuff.
None of that changes the fact that daisy-chaining is pure crap in the 21st Century. Intel made a HUGE mistake there. It's like they wanted the format to fail from day one between that and the exclusive deal with Apple (guaranteed to make the technology so far behind that it didn't stand a snow-ball's chance in Hell of competing with USB3. You say it's not supposed to, but I think you know that's simply not true.
Just like with USB2, Apple delayed putting the newer format in hoping the product they were backing (then Firewire, now Thunderbolt) would magically take off. But things don't magically take off except in the faerie world. If anything, they should have been making it cheap as hell even if it meant eating costs for the first few years. It's the ONLY chance Thunderbolt had. Now it's worse than Firewire. While Firewire was a niche format, Thunderbolt is a DEAD format and that's a BIG difference. Apple is the only major backer of it right now and they've ABANDONED the professional market for the most part (Oh Mac Pro how dead are thee?) so it's utterly pointless. USB3 Mark 2 will be out in a year or two and it will be the same badwidth as Thunderbolt at 1/10 the cost and 1 MILLION times the installed user base in short order making TB a sad memory and lesson of how NOT to do things and probably one of Steve Jobs' greatest blunders next to the Newton and Lisa (although Intel deserves much of the credit too).
Are you just trying to be daft? Two 10 Gbit/s channels per link designed to carry up to 12.9 Gbit/s of isochronous display data connected to a 6-port hub? The result would be comical. PCIe doesn't do hubs; you need a switch. And Thunderbolt daisy chains are not the same as traditional busses, due to the dual-channel, switched fabric architecture you're essentially creating a high speed ring.
Thunderbolt supports tiered stars just fine. At this point Apple has shipped at least 35 million hosts, so even if you could only convince 1.5% of the installed base to buy a switch, you'd be able to shift half a million units. Call up Intel, I'm sure they'll talk if you agree to contract for half a mil over the next 12 months. The bottom line is it's not viable economically, that's why it doesn't exist in the marketplace.
The current solution is clearly the best one available. You can connect Port Ridge based adapters that start at $29, existing displays that require little to no additional investment, and daisy-chainable 2-port solutions that cost $170.99 and up.
At this point I'm just going to call out you, KnighWRX, MagnusVonMagnum and AidenShaw as fostering prejudice towards an interface that you do not frequently interact with or fully comprehend. You are directly analogous to an online hate group that tries to convince others to join you in your mistrust and misgiving about something you are wary of due to ignorance. It just goes to show that even those who have every outward appearance of being rational, intelligent individuals can just as easily join the ranks of rednecks who denounce others because they are confused and fearful. So I have to ask, what is your angle? What do you have to gain by putting down the hard work of others? Do you not see that even though a technology is different and only embraced by the minority that it may still have merit and value in its own way?
For $200 more than a comparable motherboard without Thunderbolt.
Worth it
They have the P8Z77-V PRO model which is available with or without TB. The price difference is only $25.
Granted, the premium model is expensive.
But the price wasn't the point though. It was that they are already selling mobos with thunderbolt.
What makes you think it would be a few extra bucks, have you seen what Thunderbolt products with 1 port costs currently?? How about a solution that required these extra pieces to work equivalently?
Then why do you care?
Thunderbolt is more of a future replacement for Firewire than usb if anything.
This is such a contradiction, why did they bother to develop and add thunderbolt to all it's models then.
Both Firewire and Thunderbolt has benefits over usb (the Universal Serial Bus), it's not all about bandwidth it's also about real time reliability in low latency situations.
Too bad that Apple doesn't seem to care about pro market anymore.
What do many pro's ask? New MP, versatile xMac, matte screens, more ports (also legacy like fw) and less dongles with macs, more storage units by users choise (ODD taken away, nothing to replace it). How PRO is to take ethernet connection away from professional computer. Almost anybody that needs to process a lot of data, needs it. MacbookPRO isn't for pros anymore. It just tries to still ride with the impression...
Not to even consider Apple's software offerings...
Very good example is express card. Every owner of sony's videocam that uses SxS cards would like to have express card reader inside their macbook. Now they have to carry expensive and bulky card reader, which is not only slower than ec-slot, but also used to need external power.
And how Apple reacted to this need? They killed the only model, that had EC.
Who real pro would care if their laptop is few millimeters thinner? Or even better, their desktop? Very few pro's need to have their tools look sleek. Most would choose function over form.
AFAIK, Apple thinks that TB is useful for them only to sell ATBD. Otherwise they would kill TB as fast as they could. It is cutting badly their profits and too few users even need it. It would go the same way than ODD did.
Nevertheless it looks that Apple don't care so much about anything about their mac ecosystem. Why else they would sell ATBD without usb3 after a whole year when usb3 was put to macs? Or do they really think that in first year those who really need usb3, buy new macs and in second year, those who just like to have usb3 would start updating their appleDisplays?
Also, usb3 would be quite sufficient for many pro video usage also. Decklink had usb3 devices available way before nobody knew about TB. Everybody was then just waiting when Apple care to integrate usb3 to their models. They were the last computer manufacturer on the planet to do so.
Firewire lost the competition to usb, when in almost all situations fw would have been techinally better solution. Loosing the game also led to smaller quantities produced, which kept unit price high. If fw would have became the most used connection, I'd guess it would cost now the same than usb2 does.
Btw, pc users have had fw as long as mac users so there's not big difference in that. Windows didn't have target disk booting, but that's about all the difference there is. It is still very common just to take internal storage out and connect it to other computer as external drive. Only Apple has made this almost impossible task to do.
Usb3 also have many of those characteristics that made fw better over usb2. Also even usb2 had displaylink for screens and their new DL-3900 can handle dual 2560x1440 with usb3. Enough for most of us...
Also, external PCI has been available for years, usb-extenders are common and cheap and I'm very sure that usb will handle 4k like it's handling 2.5k now.
Light Peak was designed to go to 100G. Would you pay 10x more for your TB gadgets than what they already cost? I'd guess almost nobody would.
Philosofy in lightpeak was to get optical interconnection so hugely popular that cost per unit would be fraction of what it used to be. Then it would have been economical to make new gen of optical ports @ 10x speed. Now TB has dead end. Active copper cables make TB probably as exensive than optical ports, but the cost of optical ports are not going down, since they are not used. It will be amazing, if they ever get TB2 out with reasonable benefit over cost ratio.
And with daisy-chaining the all devices do not saturate the last link before computer?
Price for the switch would be ridiculous, but hey, so are all TB devices, so what's the difference?
I'd like to know how many situations there are that a pro wants to use a laptop with FC.
FC is usually used with SAN and switches, which aren't very mobile.
What this of course enables, is what they did in one of my workplaces: instead of MP, they bought iMac and TB-FC-box and now I'm looking reflections on the screen and a lot of beachball...
But PCIe is not going away from computers. Only Apple have this problem offering a nice expandable desktop. Apple has made pretty nice infinite loop to itself by not offering a xMac. People can't use cheap and fast add-on pcie-cards like with all other manufacturers, so you'll have to pay for über expensive TB-stuff or über expensive MP.
In windows desktops there is no need for TB, since you just add pcie-card which costs a fraction of what the mac ecosystem needs (TB-cables, TB-dongles, TB-enclosures, etc.). Also PCIe v3 has multiple bandwith compared to TB and will be so also in the future.
From windows desktop's point of view it is comical to watch how to use tech in macs. If you need usb3, fw800, esata, sas, scsi, FC or what ever, you just add a cheap card and you are ready to rock.
This also leads to another amazing feature that widows users are enjoying: drivers. I just can't understand why it is so hard for Apple to make working drivers and why does it take so long? Apple has best resources for this on the planet and very closed hardware ecosystem. Still their gpu, usb and hdmi drivers seems to suck all the time and corrections might never come. Maybe this has something to do with Apple's unique working culture, where everybody are expected to be genios all the time and the allocate too little resources for drivers. Like, one guy has one week to make a driver, because they are so fast and brilliant, that it couldn't need more people or time to tha one simple thing...
How would TB-switch be more comical than GbE-switch. Both behave the same: one pipe defines the max for the others.
Of course this would be the first place for TB2; from computer to switch greater speed and from thereon last gen speed, which usually is what the devices can handle.
Well, first you should get those 1.5% even to start using TB. Then maybe 1.5% of those might buy the switch. 35M*.015*.015= 7875 buyers market. Go for it!
It's irrelevant that some model exist. Relevant would be how many units they are selling. Does anybody know?
Btw, can you use only integrated gpu through TB with that mobo?
No, or it depends, all devices maintain full bandwidth between themselves. It's the main advantage of daisy chain compared to a bus topology. But, what is the alternative? Releasing Thunderbolt without the ability to use daisy chain would either require exclusive ports (one per device) or a purchase of an extra device just to get more ports.
A 3 way hub for example would lead to 1/3 of the bandwidth for all the connected devices, no way around it. For three devices A,B,C all communication between them would also by necessity have to be routed through, and involve the computer.
I think Apple sees the Thunderbolt port as a way to simplify the pro line. Need an Ethernet port? Get a $29 adapter and put it in one of the ports. Firewire? Do the same. It's part of Apple's philosophy (and has been) to push the envelope when it comes to dropping legacy support. That's probably why "pro" users get annoyed with them, but it is nothing new. Perhaps they see Thunderbolt as a way to address some of those issues by putting the onus on device and accessory manufacturers.
As for USB 3.0 support on the ATD, let's see what the next version brings. I agree it is a little unusual that they waited a year.
I already said Thunderbolt itself is utterly dead due to its high prices. It won't even catch on as much as Firewire did. FW800 had a clear advantage over USB2 for a long time. Thunderbolt is a solution in search of a problem. One cable sounds nice, but in practice it's too expensive and daisy-chaining is still a turd no matter how much it's polished.
About your opinion? I don't.
Since Firewire was never popular to begin with, that explains why it's going NOWHERE and why prices are through the stratosphere. Firewire had a replacement approved. It was called Firewire 3200 and it's maximum throughput is right about where USB3's usable output is and thus they deemed it would also fail and so there was no point implementing it. Sadly, I guess Intel didn't get that message.
You're trying to make sense out of Steve Jobs when you say that. Steve was an obsessive compulsive. In other words, if it appealed to him, he did it even if it wasn't good logic (e.g. look at the stall for USB 2.0 and again 3.0 all because Steve thought what he liked better should win; too bad the real world doesn't work that way. He did it again with the 22-pin iPod connector after ditching the Firewire one. He just loved non-standard crap because he felt it gave him control over the people buying it (hence the mono-polar App Store, ditching clones, etc.)
Besides, Light Peak using optical sounded really good. But once again, PRICE dictated that the theoretical prototype simply wasn't a viable market item at this point in time. By the time that it is viable, USB 4 will probably be out. You can't build a cost-efficient fusion reactor in 2013 either (there is a prototype of that being finished in France now, though. God knows when it will be a viable energy source for mass production).
It was quite clear Steve knew the future was consumer gear and the lack of support for the Mac Pro, the dumbing down of the Macbook Pro to consumer specs while retaining the name for mere show, the slow updates of Mac hardware and features for it, the lack of progress in OSX and delays because of the iPhone that persist unto this day with OpenGL still lagging behind, GPUs horribly out of date, the dumbing down of Final Cut Pro into Final Cut X when it wasn't even close to being a professional replacement when it was released, no Logic Pro update for years and no Blu-Ray support ever and it all says, "We don't care what YOU want!" and it says that to professional more than anyone.
Steve saw Lightpeak as an opportunity to get a leg up on the rest of the world and hence the exclusive deal. Sadly, he didn't get that it sounds better on paper than it does in reality and the high prices and low market share should be obvious to even a business student in college, but Steve was blind to crap like that. He's the idea guy. He had a great streak with ideas and was fortunate enough that his common sense didn't kick him in the arse more often than it did. Bill Gates was his polar opposite. Bill was a shrewd business man, not a tech guy or an innovative dreamer. Rip-off CPM and make a deal with IBM. BAM! We all know MS-Dos sucked to high heaven compared to the Mac or Amiga or even the Atari ST, but Microsoft knew business better than Commodore, Atari and Apple put together in those years.
I haven't seen any evidence that USB 3.0 has a low-latency issue with any particular hardware yet. USB 3.0's latency isn't completely fixed. A device can request a lower latency to a point and the throughput is obviously high. But until we see some actual devices, it's largely moot.
USB 3.0 Audio interfaces up until this point have been MIA and I'm still waiting for a TB one for that matter. It's largely moot as FW800 is mostly overkill as it is and that's why most audio interfaces are FW400, not 800. Most projects can easily get by with a FW400 device. Who the heck has more than 32 simultaneous audio (not counting internal midi/synth/drum) tracks playing at once for a given song anyway? Theoretically, it could handle 40 tracks. FW800 is double that. I can't imagine the song that would need upwards of 80 tracks playing at once. Thunderbolt is pure OVERKILL. That would be fine, though if it weren't so darn expensive.
Either way, I have a FW800/400 port on my Mini (my MBP has a dedicated FW400 port in addition to the 800/400 one). My own interface is only FW400 and despite some of my songs using over twenty five tracks, the FW interface was moot seeing I was recording one track at a time.
You can tell me all day long how Thunderbolt isn't meant to compete with USB 3.0, but in reality it does compete (i.e. we see hard drives for sale for both and thus market overlap) and Thunderbolt lost. Now I can see why someone who bought a 2011 Mac with TB only would want a hub that adds USB 3.0 functionality through Thunderbolt, but even that has yet to actually emerge beyond prototypes from what I've seen and the cost may not be worth it to those people when it does emerge. If it costs $400, why not put that towards a newer computer instead in many cases?
For the computer or the user, it doesn't matter what the bandwidth between the devices is. Nevertheless you only have the max bandwidth (from the link from your computer) shared to all your devices together. So for the effective bandwidth, there's no difference between daisy chain or star topology.
The difference comes, if there ever will be TB.v2 and then the star topology will be the winner. Just like with ethernet.
And if you need 3 ports and your mac only has 2?
Have you seen daisy-chainable ethernet port of fw?
For 3 ports you need one of these über expensive hubs, that you still can't even buy yet. You call this progress?
How many people really need that, though? Apple is not one to keep legacy ports around "just in case" someone needs them, and certainly not one to clutter up a 13" or 15" notebook with them. That's been the issue on the Windows side. Because manufacturers kept on shipping PCs with old ports, accessory manufacturers kept on making products for those old ports. USB improved the situation a lot, but even today Windows notebooks often still have VGA ports and others that should have been consigned to the dustbin by now. Heck, the HP Elitebook line still ships with RJ-11 ports, in case you need a dial-up connection. That's finally starting to change with Ultrabooks.
If more native Thunderbolt devices were introduced, there would be less need for them. Also, USB 3.0 adds some options, as well. You can get a USB 3.0 hub or adapter to handle Ethernet.
How many pros would like to have ethernet AND firewire AND displayport simultaneously?
How outrageous demands!
Seriously, TB is looking more and more like a "just in case port", that Apple should get rid of.
Or maybe Apple should realize, that taking away ports is not the ultimate goal for computers...