Further to that, if they had squeezed in NEC usb chipset, they'd have probably lost out on valuable PCIe lanes that thunderbolt no doubt needs. ( I think it needs at least 4x lanes, possibly even 8x )
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 to usb 3.0
- Thread starter Ryan Of Seattle
- Start date
- Sort by reaction score
Further to that, if they had squeezed in NEC usb chipset, they'd have probably lost out on valuable PCIe lanes that thunderbolt no doubt needs. ( I think it needs at least 4x lanes, possibly even 8x )
How? TB is 10Gbit in real speed terms, USB 3.0 is 480Mbps, which is 1/20th of the speed. Ergo, TB powers past USB 3.0, and there wouldn't be any degradation what-so-ever.
USB 3.0 is specified at up to 5 Gbps
USB 2.0 is specified at up to 480 Mbps
Thunderbolt is specified at up to 10 Gbps
USB 3.0 is by no means as slow as what was stated above, but rather, approximately half the bandwidth as the new TB. That said, I'm not sold on TB vs. USB 3.0, but if I had to bet, I would say that TB will take over the role that firewire currently holds, and USB 3.0 will likely still be the ubiquitous external I/O connector on computers over the next 5 to 10 years.
To the original poster: I'm sure we'll see adapters like that soon. Intel demoed a device that acted as a USB hub connected via TB.
The potential bottleneck is the adapter itself. Signal transduction takes time, but the question is how much.
Guys… shouldn't we all be in support of better I/O? Yes, USB is widespread. Yes, USB 3 is fast enough for external storage interconnects today. USB 3 is also far and away inferior in every technical respect. I, for one, very much hope that Thunderbolt gains widespread adoption, and a diverse peripherals ecosystem. I also hope the transition to photons is seamless, painless, and soon.
We would all benefit from having fewer types of ports on our computers. Faster, lower-latency, lower-level interconnects, and overall greater simplicity.
To answer the original question. It is quite possible to build an external version of anything you can hook into PCI Express using Thunderbolt. USB 3, eSATA, Firewire, 10Gbps Ethernet, DisplayPort, digital audio, an external GPU (in theory), custom solutions for specific vendors even.
Also, those of you comparing Firewire vs USB to Thunderbolt vs USB don't have your heads on straight. Apple made that mistake once, they aren't about to make it again. If Intel puts Thunderbolt on their Ivy Bridge chipset, and not USB 3… they can drive the market. Will they do that? Probably not, we don't know what the future holds. We don't know what Intel's long-term strategy is.
What we do know is this: Thunderbolt is better than USB 3 today. It is an open and diverse standard with a bright future (pun intended) that anybody can adopt today. It has the marketing and market force of today's Apple and Intel behind it (these are not niche players, if anybody has been keeping score). Apple's adoption is assured, we know that based on today's movement.
We will know more after March 2nd, iPad with a Thunderbolt port would be huge. I am pitching my lot in with Thunderbolt. It will be the interconnect of the future, and USB will not last. But call me crazy. After all, last time Apple supported a new universal Intel I/O, it never really caught on. I mean nobody uses USB these days.
We would all benefit from having fewer types of ports on our computers. Faster, lower-latency, lower-level interconnects, and overall greater simplicity.
To answer the original question. It is quite possible to build an external version of anything you can hook into PCI Express using Thunderbolt. USB 3, eSATA, Firewire, 10Gbps Ethernet, DisplayPort, digital audio, an external GPU (in theory), custom solutions for specific vendors even.
Also, those of you comparing Firewire vs USB to Thunderbolt vs USB don't have your heads on straight. Apple made that mistake once, they aren't about to make it again. If Intel puts Thunderbolt on their Ivy Bridge chipset, and not USB 3… they can drive the market. Will they do that? Probably not, we don't know what the future holds. We don't know what Intel's long-term strategy is.
What we do know is this: Thunderbolt is better than USB 3 today. It is an open and diverse standard with a bright future (pun intended) that anybody can adopt today. It has the marketing and market force of today's Apple and Intel behind it (these are not niche players, if anybody has been keeping score). Apple's adoption is assured, we know that based on today's movement.
We will know more after March 2nd, iPad with a Thunderbolt port would be huge. I am pitching my lot in with Thunderbolt. It will be the interconnect of the future, and USB will not last. But call me crazy. After all, last time Apple supported a new universal Intel I/O, it never really caught on. I mean nobody uses USB these days.
Last edited:
My bad, got 2.0 and 3.0 speeds mixed. Even so, the 5Gbps of USB 3.0 doesn't take in to account the CPU overhead, which from what I can see brings it down to 3.2. TB is still over 3x faster.
Nobody claimed to be a media professional. What are you proposing the OP and the rest of us use? There is only 1 thunderbolt product on the market right now. FW800 is slower than USB3 so there is no point in that anymore. Are you suggesting everyone just puts their computer usage on hold until more thunderbolt peripherals come out?
Very few (if any) of the media folks I work with are using USB3 (for many reasons, some stated in this thread). The primary reason right now, barring all the technical ones, is that it is not a proven tech yet. Media companies move slowly, though they are not as monolithic as large enterprise. Investing in a new type of storage will require tens of thousands of dollars (minimum). It will be a large task to set it up, is the gain in throughput and versatility worth the time lost to having to work through hiccups in the new system, and so forth.
The current popular (only?) choice for MacBook Pros is eSATA via ExpressCard. While throughput is capped by the lane width of the ExpressCard slot to some degree. The bandwidth is there to do most things one wants to do on the road. Apple's move to pull ExpressCard from the 15" was very irritating. The 17" is much less portable. Regardless, many have bought it because it was the only game in town (with murmurings of leaving Apple for another platform, I assure you).
So while the consumer market adopts USB 3 happily, the media professionals have for the most part not opted to move away from their present solutions yet (comprised of anything from fiber cards with a SAS media server to eSATA raid arrays). Thunderbolt is an interface that is VERY appealing to them. I have had numerous discussions just today with people in the business who are eyeing it with hope/lust. Obviously it is an interface that will live or die by the size of the peripheral ecosystem created for it. I think there is good reason to expect adoption in the professional world very rapidly. What that means for the consumer market and USB 3, I couldn't tell you.
Last edited:
Am I missing something.
If I plug in my 23" Cinema or 24" LCD to the display port, which I do daily, then I no-longer have this super-fast port for my super-fast hard drive I've just bought. Whoops.
I can think the only answer a genius will give me is "buy all new hardware"
If I plug in my 23" Cinema or 24" LCD to the display port, which I do daily, then I no-longer have this super-fast port for my super-fast hard drive I've just bought. Whoops.
I can think the only answer a genius will give me is "buy all new hardware"
You're missing the fact that the new super-fast HD you bought will have 2 ports, so you plug the display in to the HD, and the HD in to the Mac, and it all works.
You need to include your target market in your trite comment, otherwise your post is completely useless. Nobody uses USB for professional audio and video work. Nobody. Sure, you can fire up google and find a small handful of pro AV equipment that uses USB. But that's kiddie stuff. The vast majority is all Firewire.
If we're talking about home consumers, of course they're all using USB. End user consumers always go with whatever is the absolute cheapest possible solution to their problem. It doesn't matter that Firewire or Light Peak or anything else is faster and better than USB - if they can meet their needs with the absolute cheapest solution (USB) they will do that every time. 100%.
Ok, I'm all for TB and hope that it does indeed supercede USB, but sorry, saying no one in the world uses USB for professional audio and video work is absolute horse ****.
Composers and producers all over the world use midi keyboard controllers via USB just to name a prime example. Secondly, ever heard of an iLok? Yea, it's a USB license key that stores all of the licenses of legal sample libraries. As an owner of East West orchestral libraries, my iLok needs to be attached at all times to operate said libraries. I guarantee you East West is not "kiddy stuff".
What the frak are you smoking?
You are talking about a totally different thing. MIDI, most home audio interfaces, and iLok are all very much low-bandwidth devices. It's not like USB2 is going away... but what we're talking about is storage bandwidth. I/O intensive usages.
I wonder when we will see the first Thunderstuff.
like tomorrow? next week? in a month?
The Question I guess is how much advance info the companies got.
Does anyone know if stuff is already made or in the making?
Is this the right thread for this?
Help me
like tomorrow? next week? in a month?
The Question I guess is how much advance info the companies got.
Does anyone know if stuff is already made or in the making?
Is this the right thread for this?
Help me
What for? Are you joking? How about - to be able to have fast connection to external hard drive. As of today, Thunderbolt does not allow for this (no devices available and those that will come first will be expensive SSDs which makes little sense for external hard drive)
Pace is going to have to upgrade to any new tech if they want to stay in business. They are not the most beloved company anyway. But in their defense, their product is by nature going to annoy the heck out of people.
Dongles are a drag (literally)
No I am not talking about a completely different thing. Whether you want to talk about storage I/O or low bandith devices, it's irrelevant. The discussion here is about USB, period. And to reiterate what Dime said in his previous post:
Which is completely false. That's all I'm saying. Tons of media professionals use USB, regardless of what type of device it is. So he cannot say with 100% absolute certainty that no professionals use USB. It's a ridiculous statement to begin with.
Yea, tell me about it. I really hate dongles. I wish they'd do away with them completely.
I don't think anybody is silly enough to believe that USB is going away. Computers are going to come with USB ports for a long time to come. I like it that way, USB is a great I/O for many things. Peripherals mainly. Keyboard, mice, printers, tablets, dongles? (I hate dongles)… you get the idea. None of these are high-bandwidth devices. The argument *I* was making was USB3 vs Thunderbolt. USB3 simply isn't necessary if we have a better high-bandwidth interconnect. I would love a machine with five USB2 ports and three Thunderbolt ports. That would be fantastic. I will better qualify my statement: Professionals do not use USB3 as a high-bandwidth I/O because it has many negative traits in that application.
Hear hear.
Agree with the USB 3 vs TB argument. I am actually quite interested to see if and when we'll start to see 3rd party device manufacturer's start adopting TB in a dual port fashion, to assist in some nice daisy chaining as well.
history repeats itself
History repeats itself - again. Many of you crybabies use the same arguments that people used when the first computer manufacturers stopped shipping PCs without 2 Serial (COM) and 1 Parallel (LPT) port, shipping only USB-1 ports. People used to connect their external Hard Disk Drives via this parallel port. They made investments in this and many wanted to hold on to this technology that was tried and tested.
Today we can observe the same arguments of headless crazy USB chickens running around in confusion because of the TB disruption...
The better is always the enemy of the good. TB is better than USB-3, but does not have the market penetration yet.
(USB-1 was better than LPT/COM, but did not have the market penetration in the beginning)
History repeats itself - again. Many of you crybabies use the same arguments that people used when the first computer manufacturers stopped shipping PCs without 2 Serial (COM) and 1 Parallel (LPT) port, shipping only USB-1 ports. People used to connect their external Hard Disk Drives via this parallel port. They made investments in this and many wanted to hold on to this technology that was tried and tested.
Today we can observe the same arguments of headless crazy USB chickens running around in confusion because of the TB disruption...
The better is always the enemy of the good. TB is better than USB-3, but does not have the market penetration yet.
(USB-1 was better than LPT/COM, but did not have the market penetration in the beginning)
External storage solutions that can reach read/write speeds anywhere near thunderbolt's max seq transfer speed of 1.25GB/s (10Gb/s divided by 8) are few and far between. Although there are more storage devices with read/write speeds in the range of USB 3.0's max transfer speed of 625 MB/s (5Gb/s divided by 8), you would still need an SSD Raid or very large and fast HDD RAID.
In the vast majority of cases working with external storage, there would be effectively no difference in transfer speed working USB 3.0 or Thunderbolt, since the storage drives cannot read/write as fast as the maximum transfer speeds. The Promise Pegasus 6-bay RAID used in the engadget video was most likely in Raid0 configuration, while the safer Raid5 would not have reached the demonstrated transfer speeds of 750-800 MB/s.
We receive assets on a lot of third party external drives of varying quality, most are 7200RPM, some with eSATA, firewire, USB 3.0. Having a TB port in our workstations with easily swappable adapters for eSATA or USB 3.0 would be very helpful indeed.
It's not quite that simple. While generally one drive array will not be pushing the bandwidth of Thunerbolt. What about several drive arrays? What about multiple devices on the bus? What about USB's latency and controller model? CPU overhead? There are a lot of reasons USB is not an ideal interconnect for demanding high-bandwidth applications.