Yes, just like Firewire. And look how well that went! oh wait...
FireWire actually went very well for a long time, especially on the Mac side. It's not our fault you're new to the Mac bandwagon.
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 Coming to PCs in April 2012
- Thread starter MacRumors
- Start date
- Sort by reaction score
FireWire actually went very well for a long time, especially on the Mac side. It's not our fault you're new to the Mac bandwagon.
Yes, just like Firewire. And look how well that went! oh wait...
the thing is TB is a significantly faster than just twice as fast as usb. too fast to be ignored this time.
There's not a lot of profit in having an ultra-fast interface that isn't widely used, and/or is even a bit more expensive than the alternatives.
Firewire is a fringe technology, in large part because it's always had a limited marketplace, and never got over the cost disadvantages of selling into a small market.
USB 2.0 is a great standard. It's kinda pokey compared to Firewire, but it's absolutely ubiquitous, and it's more than 'fast enough' for most storage/data-transfer uses. The only things it's really not up for are big backups, video editing, and other 'pro' tasks where it's not enough to just do the job, but where there's value in doing it quickly - and enough value that the customer is willing to pay for the speed.
USB 3.0 covers an awful lot of the 'pro' tasks in a backwards-compatible interface requiring no changes from what you're doing today. It's fast enough that in real-world situations no rotating hard drive is going to saturate the interface, and most SSD's won't either. It's a great example of a 98% 'good enough' solution. It's cheap, fast, and effective. It's already 'won' the battle to the point where there is no battle. Apple's insane for not getting behind this.
ThunderBolt is a great Next-Generation standard/interface. Ultra-fast, data, video, SSD arrays, keyboards, Firewire adapters, PCI video card external boxes, whatever. Yay Thunderbolt. But it's so fast and so advanced that it actually requires a re-thinking of how computers and peripherals exist. It's apparently complex/expensive enough that there's only ONE company in the world that makes the freakin' CABLE, and they charge $50 for it. While it's chock-full of promise for the eventual future, there's FOUR peripherals for it, and not one of them is a basic external enclosure.
ThunderBolt is great, and it's probably the connector/standard of the Future. But USB 3 is today's standard, and Apple needs to get on the freakin' train.
Firewire is a fringe technology, in large part because it's always had a limited marketplace, and never got over the cost disadvantages of selling into a small market.
USB 2.0 is a great standard. It's kinda pokey compared to Firewire, but it's absolutely ubiquitous, and it's more than 'fast enough' for most storage/data-transfer uses. The only things it's really not up for are big backups, video editing, and other 'pro' tasks where it's not enough to just do the job, but where there's value in doing it quickly - and enough value that the customer is willing to pay for the speed.
USB 3.0 covers an awful lot of the 'pro' tasks in a backwards-compatible interface requiring no changes from what you're doing today. It's fast enough that in real-world situations no rotating hard drive is going to saturate the interface, and most SSD's won't either. It's a great example of a 98% 'good enough' solution. It's cheap, fast, and effective. It's already 'won' the battle to the point where there is no battle. Apple's insane for not getting behind this.
ThunderBolt is a great Next-Generation standard/interface. Ultra-fast, data, video, SSD arrays, keyboards, Firewire adapters, PCI video card external boxes, whatever. Yay Thunderbolt. But it's so fast and so advanced that it actually requires a re-thinking of how computers and peripherals exist. It's apparently complex/expensive enough that there's only ONE company in the world that makes the freakin' CABLE, and they charge $50 for it. While it's chock-full of promise for the eventual future, there's FOUR peripherals for it, and not one of them is a basic external enclosure.
ThunderBolt is great, and it's probably the connector/standard of the Future. But USB 3 is today's standard, and Apple needs to get on the freakin' train.
BTW
Why can't they all just replace USB 2.0 with USB 3.0 ?????
What is the point in keeping with 2.0 ????
3.0 is backwards compatible.
Stupid Apple
Why can't they all just replace USB 2.0 with USB 3.0 ?????
What is the point in keeping with 2.0 ????
3.0 is backwards compatible.
Stupid Apple
But he was responding to a post clearly discussing the idea of using both USB3 and Thunderbolt, in a topic about Thunderbolt coming to PCs. Yes, Firewire did quite well on the Mac, but on the PC side? Definitely not. Firewire is faster than USB2 and still did not take over the PC world and here we have a similar situation. Thunderbolt is much faster, but also incredibly pricey and for the majority of users the extra speed and functionality probably won't justify the jump in cost.
Last edited:
Driving technology is always good as the tech trickles down to the masses, but while I see 100Gb/s as useful for businesses I don't see it as useful for typical consumers. I see the real benefit of TB and USB3 as being a consolidation of multiple ports into a pair of ports which makes things easier for consumers.
"640K ought to be enough for anybody." - Bill Gates, 1981
BTW
Why can't they all just replace USB 2.0 with USB 3.0 ?????
What is the point in keeping with 2.0 ????
3.0 is backwards compatible.
Stupid Apple
They will also look even more stupid if the rest of the industry keep CripplePeak in the USB form. Instead of having an all-in-one connector Apple decided people needed another form of connection.
Yes, but they include
* a high-end display/docking station which provides picture plus additional USB, Firewire and Ethernet ports over a single cable,
* a high-end RAID array that ca take advantage of the bandwidth
* an ExpressCard adapter
* Pro video devices which previously required a PCI or ExpressCard slot
...none of which would have been possible with USB 3 or Firewire.
Anyway, if it doesn't take off, so what? It hasn't cost Mac users anything, simply added a second function to the existing Mini-DisplayPort. Supporting TB will be more commercially attractive after its been standard in Macs for a year or two and there are more TB-compatible machines out there.
Perhaps you've forgotten how long the original USB ports sat around, unused, on PCs before they took off.
Anyway, as lots of people have said already, TB isn't a USB replacement (a daisy chain system is no use for that). You're not going to see TB mice, thumb drives, entry-level hard drives, lava lamps any time soon, but you might be plugging them into a TB-driven hub.
Intel mobile chipsets with USB3 as standard, so they don't have to cram another contoller on the MacBook Pro/Air logic boards?
Until then, there's always FireWire (maybe not quite as fast as USB3, but it'll do for backup drives, and many Mac users have an "investment" in FireWire).
A fair point, but a (4 x SSD) RAID 0 configuration isn't particularly common .. or cheap, which isn't within the general grasp of the predominantly enthusiast reader of MR. Its impossible to say what is 'more typical', but probably a 1xSSD for a scratch disk, maybe a 4 x HDD RAID 5 are reasonable enthusiast-centric (and low end small business) approaches. For either of these, USB2 and FW800 are both already very thoroughly capped (as is also SATA-1), as well as probably USB3 and SATA-2.
It is probably reasonable to say that TB bandwidth will be more than adequate for most folks, until such time that the fiber cables come into the marketplace (IMO 2-3 years)...and for the bleeding edge adopters (of 4 x SSD RAID 0 fame) for whom there's never enough, TB is basically the best out there under the 'any price' criteria.
In the meantime, even a "mere" 1.25 GB/sec theoretically means that I should be able to download a high speed 16GB CF card in <15 seconds ... it still begs the question as to what is the I/O bandwidth performance level of that data's destination? Doesn't do me much good if it is pointed at a slow 2.5" HDD laptop drive. The lesson is that system throughput is only going to be as good as the weakest (slowest) link, and a holistic view is required.
And also timing, which has been a problem for USB for things like video. The Pros stayed with FW, whereas on the consumer level, the general approach being used by USB3 has been the "bigger hammer" bandwidth paradigm where the assumption was that the extra bandwidth will prevent time-sensitive data from arriving too late to screw things up. Works adequately on short clips.
True, but there's still going to be a competition, regardless of technical merits and arguments. It is simple human nature...if not also the desire of the USB Standards organization to try to marginalize any possible competition, be it from FW, eSATA or TB.
Yup, backwards compatibility is the basic strength of USB3. An obvious point, but one that still must be acknowledged.
Parties who need something better than USB3's limitations ... such as time-sensitive transfers that don't rely on "luck" (bandwidth overkill assumptions).
As per Wiki, it took more than a year for the first 4 USB3 devices to ship, so in comparison to USB3, it can't be that TB is an adoption laggard...so just exactly what is your attempted point?
(Waiving the magic wand)
Sonnet's TB - EC34 adaptor has shipped.
Details can be found here. It is compatible with many of their existing EC34 cards, so through it, a TB port can be used today with: E34, eSATA, FW, CF, SDXC, Aja io Express, Magma Expressbox 1, RME Hammerfall, and probably others.
I must have missed it in this article: please point out where it said that there's a law that says that Windows PCs must always be the first adopters of all new hardware technologies. A reasonable observation; HDDs are going to be an interesting technology area to monitor. My thoughts are that conventional 'spinning disk' media is still going to be with us for the next few years because it has been commoditized and as such, its cost per GB is still very compellingly low...but as you point out, it is going to be pushed further and further to the side by the superior I/O performance of solid state.
I personally suspect that the watershed turning point will be when tiny SSDs for the boot drive become the standard hardware design....I'm watching for the SSD "sticks" from the MB Air to potentially show up in the next Mac Pro as an indicator.
The fiber side of TB will provide a near-ish term performance bump; what we're really going to have to ultimately see is a non-copper approach: one can only crank up the bandwidth carrier signals so high until such time that the internal resistance simply becomes unacceptable (and max lengths become dramatically shorter). For example, if one were to today build a copper cable to put a 100GHz carrier signal down through, the entire signal would have been lost (converted to heat from internal resistance) in less than a 1m cable length.
Unfortunately, we've also seen many comparisons ... such as to eSATA ... that make this same basic error. There's more to bandwidth than just data storage. TB's potential for marketplace traction is in its unification into a single KISS connector & standard.
Personally, I expect that for certain applications ... a HDD-based RAID 1 array, for example, that FW800 will continue to be the preferred alternative until such time that TB devices come down in price...afterall, a Promise R4 at $1200 is easily twice what a FW solution with 'legacy' NewerTech Maximus externals currently costs to install. Of course, the payoff is the performance, for when one's business needs it. This ultimately hearkens back to use cases and people who need speed have to pay for it, one way or another.
-hh
The real problem is, that like many T-Bolt parts - it's expensive and not yet shipping!
Attachments
By the time they stop doing so, no doubt you'll be able to get a TB-to-FW adapter to support your legacy FW devices.
Probably. If both Windows and Mac laptops are supporting TB, it would seem to be a no-brainer.
Why does everybody have their panties in a twist?
This technology is clearly not mature enough (as in inexpensive) yet for the masses.
Product introductions for new technologies always go by the same pattern:
Initial launch, super expensive and a few pioneer users will go with it.
Then it takes a while and manufacturers figure out mass production and change their product lines to include this new technology.
Sometimes a competing product is introduced sooner (or later) widespread adaptation in products occurs. (As in new PC's)
Most users, whose "old" stuff works keep running their old stuff and get the new technology when they buy new computers.
In the meantime the 2 technologies emerge and eventually all computers will have both.
Why would this be any different this time?
BTW: In this case Apple will IMO eventually have to cave in and also implement USB3, because of PCs. While they don't officially come out and say it, they want their products to be as much as possible compatible for switchers.
A lot of people think Firewire 'flopped'. It didn't. It became an industry standard and was found on devices all over the place, things that weren't for the personal computer market. Literally every video camera had it on it. It was never going to be mainstream, it wasn't designed for keyboards, mice, printers, etc. Don't forget how literally useless USB 1 was for storage. Firewire also did so many things USB can't, by design. Boot from another Mac as a host device? Have more than one device interact with another?
USB 3.0 is great. Almost all consumer storage devices are using USB 3.0. You're not going to find Thunderbolt jumpdrives or Thunderbolt hard drives for $50 at Wal-mart. Apple needs to have it on their devices. USB 2.0 is a huge limitation. Compared to Firewire the speeds were 'good enough', but compared to new technology it just isn't. Neither is FW800 and that's why Apple needs to add USB 3.0 to their computers.
Thunderbolt probably won't go far soon. It'll have a certain high-end market but that's about it. Not many people will spend $900 for an external video card. Only a small handful of devices even exist. It's not going to replace USB or HDMI, that isn't what it was designed to do.
USB 3.0 is great. Almost all consumer storage devices are using USB 3.0. You're not going to find Thunderbolt jumpdrives or Thunderbolt hard drives for $50 at Wal-mart. Apple needs to have it on their devices. USB 2.0 is a huge limitation. Compared to Firewire the speeds were 'good enough', but compared to new technology it just isn't. Neither is FW800 and that's why Apple needs to add USB 3.0 to their computers.
Thunderbolt probably won't go far soon. It'll have a certain high-end market but that's about it. Not many people will spend $900 for an external video card. Only a small handful of devices even exist. It's not going to replace USB or HDMI, that isn't what it was designed to do.
Finally! Hopefully we get some cheap TB gadgets soon
I'd rather see inexpensive TB gadgets than cheap ones since the top 3 definitions of 'cheap' in the dictionary are: contemptible, despicable, shameful.
----------
Have you priced out a simple HDMI cable lately? OMG!
Like when I bought a motherboard that had USB 1.0 (around when USB was released). By the time I actually owned a USB device, I got a new motherboard, and it had USB 1.1.
ThunderBolt only has a chance if it is integrated onto the PCH at those controller prices. The cable pricing and niche within a niche is not helping either.
USB 3.0 came in at $4-5 for the original controllers.
USB 3.0 came in at $4-5 for the original controllers.
On would think that after the previous battles that facts would be a little easier to come by.
Fact: MiniDisplayport is THE standard. The USB group very early on rejected the use of TB over a USB connector. Sony's USB LightPeak connector is an orphan.
Fact: Intel has a roadmap to 100 Gb/s over copper by adding more channels.
Fact: TB supports channel syncing, and isochronous timing protocols like FW and data channels are inherently low latency (8ns), features that would be desirable for video and audio production; an inherent advantage over USB 3.0.
Fact: TB is inherently multiprotocol.
Fact: Apple's current TB implementation will be compatible with future optical implementations; the transceivers will be imbedded in the cable. Apple's current implementation will be backwards compatible with future TB implementations.
The only speculation would be which future Ivy Bridge implementations on the various Mac's and TB displays will support both TB and USB 3.0 connectors?
I would guess all.
I would also speculate that TB will ultimately be the Mac Pro's assassin, though I hope that is far down the road.
Fact: MiniDisplayport is THE standard. The USB group very early on rejected the use of TB over a USB connector. Sony's USB LightPeak connector is an orphan.
Fact: Intel has a roadmap to 100 Gb/s over copper by adding more channels.
Fact: TB supports channel syncing, and isochronous timing protocols like FW and data channels are inherently low latency (8ns), features that would be desirable for video and audio production; an inherent advantage over USB 3.0.
Fact: TB is inherently multiprotocol.
Fact: Apple's current TB implementation will be compatible with future optical implementations; the transceivers will be imbedded in the cable. Apple's current implementation will be backwards compatible with future TB implementations.
The only speculation would be which future Ivy Bridge implementations on the various Mac's and TB displays will support both TB and USB 3.0 connectors?
I would guess all.
I would also speculate that TB will ultimately be the Mac Pro's assassin, though I hope that is far down the road.
I highly doubt TB will become mainstream - USB3 is quite suitable for the majority of the user base, and the devices are cheap. Those that really require TB ( especially for storage ) are fairly few are far between in comparison and will continue to buy FW / TB. Sure, TB is nice, one cable it may be faster, nicer technology etc etc, but at the end of the day the average user doesn't care - there's already a standard that is suitable for the majority, namely USB. Its the average consumer who decides what is 'mainstream'. By the time USB3 becomes too slow for the technology of the time, the next iteration of USB will be near / already released.
USB is for the masses, TB for the minority.
I'm surprised no one has come out with a USB dongle for TB -> USB3. Sure, a driver would need to be developed, but shouldn't be that difficult. Its a shame because personally, I'd love to buy a couple USB3 external hard disks, USB2 is too slow and USB3 drives are cheaper than FW ( although I'm absolutely happy with my existing FW drives ).
USB is for the masses, TB for the minority.
I'm surprised no one has come out with a USB dongle for TB -> USB3. Sure, a driver would need to be developed, but shouldn't be that difficult. Its a shame because personally, I'd love to buy a couple USB3 external hard disks, USB2 is too slow and USB3 drives are cheaper than FW ( although I'm absolutely happy with my existing FW drives ).
Last edited:
Maybe someone like me, who doesn't have hardware that will benefit from ThunderBlunder's speed or a wallet full of cash to burn on rare super-expensive hardware. Not to mention that virtually every transmission technology is beholden to "theoretical" maximums, including TB and FireWire before it. I guess you missed the day when they explained that a chain is only as strong as its weakest link.
Good luck with that.
The data side is 10 Gb/s in both directions, and the display side is also 10 Gb/s in both directions. So the total throughput is indeed higher with TB.
It may be pricey now. But in 5 years, I guarantee you SSDs won't be "rare" or "super-expensive". Then it will be a benefit to have the available bandwidth that TB offers, rather than sticking with USB.
Every bus has overhead, and none of them reach their "theoretical" maximums. I never denied that. However, some are more efficient than others. FW for example, comes closer to its "advertised" speeds than USB ever did.
Also, you missed my point again ... I was merely pointing out that USB bandwidth was not as large as he thought it was. The theoretical peak was not even what he thought it was.
One thing to remember, to those who say they will be able to move data from their thunderbolt drives (multiple SSD's) to the macbook pro.
This will only work if your laptop has a hard drive that can take it. It does not matter if your thunderbolt solution can write at 600 MB/s if you are using a cheap standard 5,400 rpm 2.5" hard drive to write to it. Data is never going to be read at more that the drive allows (likely 100 MB/s). The slowest link is the weak link that will determine everything. So to make use of SSD's in raid 0, you better have SSD's in raid 0 in your computer.
Thunderbolt, i predict will be a little like firewire, used only be professionals, who really need it. The average consumer will never need TB.
Those who dream of 4x graphics solutions over TB, really? over a PCI express x4 slot? That would be bottlenecked so bad....
I don't think 100 gbps is over copper, thought it was optical.
This will only work if your laptop has a hard drive that can take it. It does not matter if your thunderbolt solution can write at 600 MB/s if you are using a cheap standard 5,400 rpm 2.5" hard drive to write to it. Data is never going to be read at more that the drive allows (likely 100 MB/s). The slowest link is the weak link that will determine everything. So to make use of SSD's in raid 0, you better have SSD's in raid 0 in your computer.
Thunderbolt, i predict will be a little like firewire, used only be professionals, who really need it. The average consumer will never need TB.
Those who dream of 4x graphics solutions over TB, really? over a PCI express x4 slot? That would be bottlenecked so bad....
I don't think 100 gbps is over copper, thought it was optical.
somewhat unrelated question...
I'd like a little extra space available for my bootcamp partition on my mac air. Would a 32gb class 10 sd card be faster than an external hdd (connected via usb 2.0)? - I guess I'm asking which is a faster port - the usb 2.0 or the sd card slot?
I'd like a little extra space available for my bootcamp partition on my mac air. Would a 32gb class 10 sd card be faster than an external hdd (connected via usb 2.0)? - I guess I'm asking which is a faster port - the usb 2.0 or the sd card slot?