Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Whilst adding USB 3.0 would be a nice thing to do. I wonder if that could send a message to the market making companies think Apple is supporting USB rather than Thunderbolt. As I see it, TB is faster than USB therefore it is superior, which means we should use that and not USB 3.0. But of course there's the issue of cost.

It just seems sad to use USB 3.0 when TB with mini display is capable of so much more.
 
This whole conversation is stupid.

Of course Apple needs to go with USB3.

They HAVE to have a USB port on their machines. There's no way they can have TB only, because so many peripherals require USB.

Given that, it would be crazy to go with USB2.

The only counter-argument is that you could use a TB->USB3 adapter, but that just seems silly given the number of USB devices out today. (printers, thumbdrives, external hard drives, scanners, cameras, mp3 players, etc)

No. The real counter argument is that that al those devices are addressable using wifi or bluetooth (besides the thumb drives).

MP3 -> iTunes match/iCloud,
external hard drives: NAS/Time Capsule
Printers: Well every 100$+ printer offers a wifi card or at least an ethernet port (same with scanners)

Camera's just don't use USB. They use Firewire and SDHC.
 
I think in the end, the success of Thunderbolt comes down to whether Microsoft will add Thunderbolt port support to Windows 7 the beginning of 2012 (through a driver update) and support it natively in Windows 8.

If Windows 8 includes Thunderbolt support, then it will become a success, because the Thunderbolt port would become a real replacement for the external Serial ATA (eSATA) ports used by many external hard drives.

USB 3.0 will primarily be used for occasionally-connected external hard drives and connections to digital SLR's.
 
The point of the original article

seemed to be that although it's pretty much a foregone conclusion that Apple will include USB 3.0 support when they start shipping Ivy Bridge systems, they are also looking at including it before then.

If anyone paid attention to how little of USB 2.0's 480 Mbps was actually attainable in real world scenarios (generally low 30 MB/s and never more than 40 MB/s) they would realize how much protocol overhead USB brings to the table. USB 3.0 uses 8b/10b encoding, so the 5 Gbps nominal symbol rate is immediately cut down to 500 MB/s, and after protocol overhead it is unlikely that > 400 MB/s will ever be achieved even with just a single device connected to the bus. There are several SSDs on the market that can deliver well over 400 MB/s streaming reads and writes, and yet how do these fare when connected to a currently shipping USB 3.0 port? Anandtech just maxed out an Asmedia ASM104x USB 3.0 controller at 326 MB/s write and 303 MB/s read. That's throughput efficiency in the ballpark of 60-65%, not counting the 8b/10b encoding which is unavoidable.

Renesas (NEC) announced in March their 3rd gen USB 3.0 controllers which offer 40% better write throughput and a 90% reduction of power consumption in standby, and now we have a rumor that Apple is considering including USB 3.0 ahead of Ivy Bridge. I'm pretty sure the reason Intel and Apple have both held off on USB 3.0 is that the technology is still a long way from maturity, and the currently available silicon has some serious drawbacks. Also, Apple probably wants their driver to be polished enough that throughput won't lag noticeably compared to Windows, as was the case with some of their USB 2.0 implementations.

Thunderbolt, in its first commercial release, offers 2 channels which simultaneously provide 10 Gbps to the protocol layer, which in the case of PCIe on Sandy Bridge systems has a theoretical maximum efficiency of around 82%. Anandtech has already demonstrated actual throughput at just over 80% efficiency with currently shipping hardware. This is what's possible when you create a new architecture and don't have to worry about whether 4 billion devices produced over the past 15 years will still work when connected to it.
 
Last edited:
I think in the end, the success of Thunderbolt comes down to whether Microsoft will add Thunderbolt port support to Windows 7 the beginning of 2012 (through a driver update) and support it natively in Windows 8.

As you note, if the support can be added via a driver update there's probably no need for Microsoft to support it. Intel could write and ship the drivers. For initial installation onto a TBolt drive, the F8 escape should work (especially if the PCIe SATA controller in the TBolt device is already supported by Windows).

Windows already supports hot-plug PCIe (although PCIe x1) in the form of ExpressCard slots - it's possible that TBolt could look enough like ExpressCard that it could simply work.

Windows also supports dismounting of SATA disks and arrays, so that breaking the chain would not require a reboot.
 
Perhaps Tim doesn't carry arrounf the same emotional bagg(ages) of hurt and we might even see blueray.







Apple won't keep putting USB2 when all other machines support USB3

Apple NEVER said, there won´t be USB3 a some point (Intel´s support).
This thing has NOTHING to do with Jobs or Cook, how can people be so naive?
Btw SJ is still at Apple.
Btw nobody needs blu ray or optical media these days.
 
I would hope so. USB 3 will become a standard, regardless of thunderbolt. Thunderbolt will always be the expensive, but fast niche, like FireWire.

I enjoyed your analogy. although on paper thunderbolt is great, if they had gone USB 3.0 instead, we would all have been able to enjoy it from the start. Their are way more USB 3.0 products out and about in the market already. Its backwords compatible etc. They could have used that extra room for extra battery life or (for the MBA) a GPU which would have made the macbook air a super powerful light notebook.

Well we missed it this round, and we always need a reason to upgrade, so hopefully with the lovely ivy bridge (faster w/less power!) we can enjoy USB 3.0 and thunderbolt as well :rolleyes:
 
Why does Apple have to have USB on their machines?
Wireless mouse/keyboard/printers/scanners.

Yeah, let's just keep the options open. ;) I don't care about having to change batteries every few weeks/months in devices that really have no advantage in being wireless.

I hope this new "we're open to options" mentality at Apple translates to some of the MacRumors members. The closed mindedness some of you display is appalling.
 
How can a computer with one TB/mDP + two USB 2 ports (existing 13 and 15" MBPs) serve more 'legacy' USB 2 devices than a computer with one TB/mDP port + two USB 3 ports?

A USB 3 port does not provide more 'legacy' support than a USB 2 port.

But it does provide support for USB 3.0 devices.

Which is an advantage. :)
 
It just seems sad to use USB 3.0 when TB with mini display is capable of so much more.

I don't think TB and USB3 are really in competition.

TB is more expensive (device chips, smart cables, everything needs a daisy-chain connector) but can potentially do things that USB isn't very good at (super-fast SSD RAID arrays, external graphics cards and PCIe slots, single wire display/docking stations like the Apple Thunderbolt display).

I don't think we'll see TB memory sticks, printers, cheap backup drives any time soon (although the article did say that Apple was working with suppliers to produce cheaper external drives). On the other hand, I don't think we'll see people hanging external graphics cards (well, maybe low-performance ones) or 10GB Ethernet interfaces off USB3. TB is more a replacement for ExpressCard (or internal PCIe slots) than USB.

They could even be complementary: e.g. someone could produce a USB3 controller - maybe alongside other ports on an external dock - that could hang of TB. Heck, maybe Apple plans to stick a USB controller in future versions of its Thunderbolt display.

In the meantime, how much do Mac users really want USB3? We have FW800 which is much better than USB2, and will do me as a stop-gap until either cheaper TB drives appear or someone does a TB-to-eSATA adapter. You can get external drives with FW800 and eSATA connectors...
 
I'm sticking with USB 2.0 thank you very much! I don't really mind if my file transfer takes a bit longer. There's nothing I hate more than competing standards: you never know what to buy, whether it will last long enough, and what will happen with support in the future. For all I know, USB 2.0 is cheap and works fine. Even FireWire is hardly supported in most peripherals. I can't even find a proper FireWire external hard drive that doesn't cost a fortune. FireWire never caught on, it's just another Apple standard that tried to replace USB but never managed.

Just FYI, a USB 3.0 port will also accept USB 2.0 devices. So if a device is upgraded, it will perform several times faster (theoretically). If it is an older USB device (say from 2011!) it will still operate just fine. In other words, backwards-compatible, not competing.
 
I think in the end, the success of Thunderbolt comes down to whether Microsoft will add Thunderbolt port support to Windows 7 the beginning of 2012 (through a driver update) and support it natively in Windows 8.

If Windows 8 includes Thunderbolt support, then it will become a success, because the Thunderbolt port would become a real replacement for the external Serial ATA (eSATA) ports used by many external hard drives.

USB 3.0 will primarily be used for occasionally-connected external hard drives and connections to digital SLR's.
Hardware needs to be there as well.

where are the data only Thunderbolt cards?

where are the pci-e video cards with Thunderbolt?

where are the amd boards with Thunderbolt?
 
I'm so happy I waited for a reason. I already own four storage devices that are 3.0 and it's already gone mainstream (w/ some people don't even know they have a 3.0 on their device). It's going to be the standard whether Intel likes it or not. *whiny intel bitches*

Thuderbolt w/ probably supersede firewire and video editors may praise it, but I think for the rest of the folks they'll be sticking to format that's accepted everywhere.
 
Thunderbolt will become the "standard" peripheral interconnect interface....

Apple will drop USB eventually just like they did with MiniDIN Serial in 1998 when the iMac, Wallstreet PowerBook G3, and B&W G3 came out.

Remember, they dropped ADB too at the same time. No more Mini-DIN.

Thunderbolt is going to be niche at best. The average consumer neither wants nor needs that kind of expensive crap that really has only a couple applications.

And the B&W G3 came out in 1999 (as did the Wallstreet PB G3... which did have both serial and ADB.. and of course as we all know, the B&W G3 had an ADB port as well). As of 1998, the Beige G3 was still very much carrying serial ports... although those had been around for more than a decade, and even then no one really used them. Apples and oranges. ADB absolutely had its uses for a long time, but naturally given that it wasn't hot-pluggable, and like their serial connection, was PROPRIETARY, and not an industry standard, it was now free to go.

I find it interesting that you compare Apple's own proprietary hardware (ADB, Serial) to something that's an industry standard (USB) and claim that they're obviously going to do away with it. I think you might be guilty of being a bit of an idiot in this instance.
 
Last edited:
Where is my FREE Fiber-Optic 100 Gigabit Fiber Internet connection, Apple...Google?? Anyone??? It is is 2011 folks, and we are stuck with the same cable broadband speeds we had in 1997!!! 14 year old technology, you would think our connection speed options would have improved....

In 1997, I was dialing up to the internet at 28.8 Kbps. By 1998, 53.3 Kbps. By 2000, 1.5 Mbps. Currently 10 Mbps. Yeah, I'd say things have improved very considerably. That's about 200x faster than what most people used in 1997.

In 1997, Macs shipped with 200MHz processors,

350 MHz was top of the line, actually.

32MB of RAM

64 MB was standard on some; expandability went up to 1.5 GB. We're actually only about 10-20x as expandable as before.

and a 4GB Hard Drive, now we have computers shipping with over 200x as much RAM, are at least 50x as fast and have 500x as much disk space...

But the internet speeds are still the same????

Right. Well, #1 they aren't the same, NO ONE had cable that fast in 1997, even schools had slower cable. The extremely rich who constitute 0.1% of the population don't really count here.

You're also forgetting rather conveniently that hard drive speeds haven't increased by 500x the same way hard drive capacity has. You could get 15 MB/s out of a standard hard drive back then. We're at around 125 MB/s for an HDD, more for an SSD, but that's still relatively new, and it's far cheaper to implement HDD-based server storage. Couple that with the fact that the number of internet users has increased probably two orders of magnitude (yep, well into the billions now... then it was probably not very far into the double-digit millions), and you've got a TON more demand on those servers.

Do you have the faintest idea how much traffic the internet deals with every day? And you're suggesting we multiply that a hundred fold?

I want my 100 Gigabit Fiber Internet connection....NOW!

So pay for it.

We should be able to transfer at least 500 Megabytes of data per second over our Internet connections in major cities..[/quote

Ah yes. A hundred million people all downloading at 500 MB/s (not that that's possible, as very few people have hard drives capable of even 1/4 that speed). And all we'd need is an ISP with a peak bandwidth of 50,000,000,000 MB/s.. or you know, 50,000,000 TB/s, 50,000 PB/s. I'm sorry, how many SSDs would that be? A hundred million plus?*Well, that wouldn't cost much, would it? $25 billion, tops.
 
Which is tricky language - the "maximum throughput" is the throughput of the chain, not the maxium throughput of each TBolt device.

The picture as shown would seriously oversubscribe the bandwidth of TBolt, so that all the devices would not be able to simultaneously run at the maximum bandwidth of the device.

(Oversubscription per se is not a bad thing, unless you require every device to run at its maximum.)


I think Sjohnny provide a good answer to this - at least a start
 
Thunderbolt is going to be niche at best. The average consumer neither wants nor needs that kind of expensive crap that really has only a couple applications.

And the B&W G3 came out in 1999. As of 1998, the Beige G3 was still very much carrying serial ports... although those had been around for more than a decade, and even then no one really used them. Apples and oranges. ADB absolutely had its uses for a long time, but naturally given that it wasn't hot-pluggable, and like their serial connection, was PROPRIETARY, and not an industry standard, it was now free to go.

I find it interesting that you compare Apple's own proprietary hardware (ADB, Serial) to something that's an industry standard (USB) and claim that they're obviously going to do away with it. I think you might be guilty of being a bit of an idiot in this instance.

Apple already makes the single TB device that many consumers are going to purchase; The Apple TB Display, which should start shipping next week or so. The display masquerades as a dock with a Magsafe and TB cable to the Mac, and includes three USB 2.0 ports, a single Firewire 800 port, and a single Gigabit Ethernet port, plus a second TB port for daisy chaining.

What's not to like. Look for third parties like EISO to offer similar capabilities.
 
In 1997, I was dialing up to the internet at 28.8 Kbps. By 1998, 53.3 Kbps. By 2000, 1.5 Mbps. Currently 10 Mbps. Yeah, I'd say things have improved very considerably. That's about 200x faster than what most people used in 1997.

I 1997, I had cable broadband. 56 kbps was already quite widespread even though you had competing standards (X2 vs K56Flex, this is before V.90). You were really on the low end with your 28.8 kbps there, 33.6 was already old hat in 1997.

350 MHz was top of the line, actually.

Not in Intel land it wasn't. Deschutes were introduced in 1998, January and only were available in 333 mhz configurations at first. 350/400 mhz variants were introduced with the BX chipset and the 100 mhz FSB option.

The Power Macintosh G3 (Beige) introduced in November 1997 came with a 333 mhz processor made by Motorola as the top of the line. It was also based around the 66 mhz FSB.

350 mhz only came into view in 1998 for every player in the consumer space. In 1997, I'd say 300 mhz was top of the line for the greater part of the year.
 
Apple already makes the single TB device that many consumers are going to purchase; The Apple TB Display, which should start shipping next week or so.

I've worked in many office buildings and home offices. I see a whole lot more Dell displays than Apple displays. The few times I do see an Apple display, its almost always an iMac, not a stand alone monitor. TB is going to need a heck of a lot more than that going for it if it's going to gain widespread adoption.
 
...although on paper thunderbolt is great, if they had gone USB 3.0 instead as well, we would all have been able to enjoy it from the start.

I suggested a change in your post - it doesn't have to be either/or. Both TBolt and USB 3.0 would have been good.


Heck, maybe Apple plans to stick a USB controller in future versions of its Thunderbolt display.

Did you mean USB 3.0? It already has a USB 2.0 controller.


Hardware needs to be there as well.

where are the data only Thunderbolt cards?
where are the pci-e video cards with Thunderbolt?
where are the amd boards with Thunderbolt?

Good point - maybe we'll see these when the Mac Pro get TBolt....

----------

It's going to be the standard whether Intel likes it or not. *whiny intel bitches*

Intel uses the NEC USB 3.0 controller on most of its Sandy Bridge motherboards.

Intel isn't trying to block USB 3.0 - a much more logical explanation is that USB 3.0 wasn't stable enough when the engineering schedule for the first round of Sandy Bridge chipsets was being planned. Therefore, not in the schedule, not in the chipsets - so Intel buys the NEC controller.
 
Where is my FREE Fiber-Optic 100 Gigabit Fiber Internet connection, Apple...Google?? Anyone??? It is is 2011 folks, and we are stuck with the same cable broadband speeds we had in 1997!!! 14 year old technology, you would think our connection speed options would have improved....

In 1997, Macs shipped with 200MHz processors, 32MB of RAM and a 4GB Hard Drive, now we have computers shipping with over 200x as much RAM, are at least 50x as fast and have 500x as much disk space...

But the internet speeds are still the same????

I want my 100 Gigabit Fiber Internet connection....NOW!

We should be able to transfer at least 500 Megabytes of data per second over our Internet connections in major cities....the infrastructure is there, but it is being hoarded by institutions, companies, and Universities, and not allowed for mass-public subscriber access....our only two main options for "semi-fast" Internet are still cable and DSL....FIOS is only on Verizon in "select markets" and it is not a full Fiber connection, at that.

How do you do your math... In 1997 internet speeds were topping out at around 30 kbps (lets just say 35 kbps to be on the safe side) and that was just downstream, upload was no where near that. Today I have Fios internet connection and get a 35/35 mbps symmetrical connection (and this isn't even their top package, I can get 100/50 mbps). My speed tests actually peg me a 42mbps down and 32mbps upload.

1000 kbps = 1 mbps. So 35mpbs (my connection is about 35,000 kbps).

So if you take the 1997 number of 35 kbps and compare it to my now 35,000 kbps conneciton it is 1,000x faster. So it has actually progressed MUCH more than other computer technologies, or at least as fast. And this isn't even taking into account the MASSIVE improvements fiber optics bring in upload speeds. If you look at 1997's best offerings of 50kbps versus Fios's top offering of 100mbps, the improvements are even greater.

Let's not even get into the Fios demos of the 1gbps speeds and Google's actual network near stanford that runs on 1gbps...

And the fact is, Fiber Optic cables that Verizon has run are able to handle much faster speeds than this now. the current bottleneck are the processing points and hardware that translate the lines. These are expensive machines and to be honest, 100mpbs is overkill right now, let alone 1gbps...

And for people saying blame the crappy U.S. Government for slower U.S. speeds compared to the rest of the world - What the heck are you even talking about? The reason the U.S. has slower speeds (for the most part) is not due to government regulations or lack there of, but because the United States is one of the most spread out countries, with a fairly even population density across the continent (when compared with other countries).

So the cost network infrastructure (wireless, cellphone, cable, broadband, fiber, etc) is MUCH more expensive and takes much longer.
 
still considering? Why is it taking so long? It is clear that USB 3.0 will become standard because people are familiar with USB. And I'm sorry, but Thunderbolt will not take off, I understand the investment they made into it, but they have to realize that thunderbolt is dead, it was since day 1. Outside of tech geeks, average consumer doesn't even know what thunderbolt is, but they do know USB. Thunderbolt is firewire all over again.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.