New Thunderbolt Revision Features 20 Gbps Throughput, 4K Video Support

hooligun writes "The next-gen Thunderbolt tech (code-named Falcon Ridge) enables 4K video file transfer and display simultaneously in addition to running at 20 Gbps. It will be backward-compatible with previous-gen Thunderbolt cables and connectors, and production is set to ramp up in 2014. An on-stage demo with fresh-off-the-press silicon showed the new Thunderbolt running 1,200 Mbps, which is certainly a step up from what's currently on the market."

Comments

Adoption by Mass Market?

[AtomicSymphonic]

So, will we see OEM Windows PCs come by default with Thunderbolt ports? Or is this another fantastic, magical, extraordinary Apple Inc. exclusive?

Re:Adoption by Mass Market?

[OhANameWhatName]

So, will we see OEM Windows PCs come by default with Thunderbolt ports? Or is this another fantastic, magical, extraordinary Apple Inc. exclusive?

You wouldn't seriously risk upgrading to Windows 8 just to be able to use 20 Gbps external connections would you???

Re:Adoption by Mass Market?

[Radagast]

Asus has a motherboard.

Re: Adoption by Mass Market?

[UnknowingFool]

I don't know what you've been smoking but Thunderbolt is an Intel invention. They worked with Apple on implementation with Apple's most obvious contribution being the VESA compliant mini-Displau port connector. For their efforts, Apple got a good six month lead on their competition as they had products the day Intel released the specs. Incidentally, Apple got the Thunderbolt trademark and then transferred it to Intel.

Re: Adoption by Mass Market?

[Anonymous Coward]

its*

Re:Adoption by Mass Market?

[Guppy06]

Would even that be enough? I have a motherboard with Firewire ports, but the only use I get out of them is when I need to connect my sister's MacBook for some reason.

I've got more USB 3.0 devices than I do Firewire.

Re: Adoption by Mass Market?

[UnknowingFool]

Really? So Apple's marketing is sompowerul that Intel's website is spewing Apple's propaganda? Or change Wikipedia:

Thunderbolt (codenamed Light Peak)[1] is a hardware interface that allows for the connection of external peripherals to a computer. It uses the same connector as Mini DisplayPort (MDP). It was released in its finished state on February 24, 2011.[2]

On the same day, Apple released new iMacs with Thunderbolt.

Now what are your facts?

Re:Adoption by Mass Market?

[DigiShaman]

Figured it would be the P8Z77 lineup. With USB3, I honestly did see the need for thunderbolt so soon. I just didn't see the need for slinging that much data back and forth. After experiencing the first generation of USB1, it wasn't used much at all (if ever). This is how I see Thunderbolt now. So I opted for P8Z77-V Pro instead.

Re: Adoption by Mass Market?

[Anonymous Coward]

[citation needed]
What exactly are the poor engineering choices for which Apple is pushing this stuff? Why is the overall value and appeal dubious? I only ask because somehow you're at +4 Interesting without having actually said anything.

Re:Adoption by Mass Market?

[Chrontius]

A counter-example: Last time I bought a desktop computer, I went out of my way to buy one with Firewire(400, at the time, was cutting edge). I then proceeded to buy an external hard drive for routine backups, and later daisy-chained an iPod 4g off it. It beat the pants off the USB cable that the iPod came with, at the time.

I had no idea what I was going to do with it at the time, though.

This time, I went with one of the Gigabyte thunderbolt motherboards. God only knows what I'm going to do with it, but when I figure it out, it's right there. Maybe I'll add a Drobo with thunderbolt and gigabit ethernet, when they introduce one. Maybe the next generation of graphics cards will be external, to isolate it from CPU and PSU heat, or because they run much cooler with a dedicated PSU due to voltage concerns, or a push-pull fan configuration, or because they're immersed in a sealed oil tank and there's nowhere in a conventional case to mount it.

Point is, future-proofing, followed by an incremental upgrade, is often cheaper than buying two or three mid- or low-end systems over the course of the computer's life. You have to be willing and able to do the upgrades, but without the interesting new busses to connect them with, you may not be able to do it at all.

Re: Adoption by Mass Market?

[Anonymous Coward]

Perhaps he's referencing the fact that it's still rare to see systems that support Thunderbolt.

Re:Adoption by Mass Market?

[jtownatpunk.net]

I used to use firewire all the time back when I used to do a lot of video editing around the turn of the millennium. The first generation of USB was so bad that I didn't even consider USB2 for my external storage. Firewire, OTOH, was a rock. Never had a device just disappear for no reason. Throughput was better, CPU load was lower, isochronous transfer was possible. Night and day. Like comparing a Lexus to a Yugo.

Of course, now all my stuff is USB because firewire components are so rare and I have no need to move devices between computers. I've got gigabit ethernet to move files and I don't need to move a single optical drive between multiple machines. And USB is much more reliable than it used to be. My new gaming rig has two firewire ports but I haven't used them. Neither of my laptops has a firewire port and I haven't missed them. Thunderbolt seems like a solution to a problem that no longer exists [in my world].

Re:Adoption by Mass Market?

[cheater512]

Your saying Firewire was great because you could plug a whopping 2 devices in to it? And both those devices had USB connectors which everything else used?

I'm not saying that Firewire isn't technically better than USB for several things.
Its just a poor reason to say its great because you managed to find something to plug in to it.
FYI I have never used my firewire ports - nothing I have uses it.

Is a separate technology really required just for hard drives? Not really which is why USB 'won'.

Re:Adoption by Mass Market?

[noh8rz10]

there was a rumor today that Apple is going to release an update to the Mac Pro this month. But imho the mac pro is obsolete and has been replaced by a macbook with a thunderbolt port. This covers all the I/Os that you would normally need a big rig for. the only exception I can think of is upgrading the graphics card down the line. But I don't know enough about the state of the graphics card industry to say if the top-of-the-line discrete card in the macbook pro is not good enough for desktop uses.

Re: Adoption by Mass Market?

[Anonymous Coward]

note: jedidiah is a gnu/hippie who's angry that Apple took-over the *nix desktop market.

In any case, Thunderbolt has been out for two years now and the peripheral selection is pretty pathetic. Apple has an expensive monitor/docking station. Belkin's docking station has been "coming soon" forever now. And there's some drive enclosures, and that's abou tit.

Re:Adoption by Mass Market?

[Anonymous Coward]

Firewire, OTOH, was a rock. Never had a device just disappear for no reason.

Mac yes, Windows no. I used Firewire setups for years and drives were always dropping, logs filling up with "sbp2" and "ntfs" errors, etc. (This was with 3 different controllers and lots of drives.)

I used Firewire because it was "better", but to be honest, for the normal random-access use case there's no noticeable difference with USB2.

Re:Adoption by Mass Market?

[Slashcrunch]

Magical and exclusive? Do you mean like when Windows PCs started to ship with FireWire ports?

Re:Adoption by Mass Market?

[smash]

It's not a mass market consumer technology. Itsa the modern day equivalent of SCSI.

Re: Adoption by Mass Market?

[PhunkySchtuff]

There's some really good storage available with Thunderbolt now. I can get an 8-bay RAID enclosure with Thunderbolt for around a grand (bare enclosure with no drives) put 8 drive mechanisms in it and get a multi-terabyte array that delivers around 650MB/sec (megabytes, not megabits) per second read and write on my MacBook Pro.
Prior to this you needed really expensive FibreChannel equipment to deliver the same kind of performance.

Re:Adoption by Mass Market?

[Anonymous Coward]

Disagree. Thunderbolt is only interesting with laptops & low-profile desktops (e.g. Apple kit). Everyone else already has "SCSI" (SAS/SATA) and a full compliment of ports.

Except at this point, the docking/expansion options aren't really there yet.

Re: Adoption by Mass Market?

[Mad Merlin]

Prior to this you needed really expensive FibreChannel equipment to deliver the same kind of performance.

No, not really. You can get an 8 bay enclosure (like this) with SFF-8088 connectivity for half a grand and get 4 gigabytes per second read/write.

Re: Adoption by Mass Market?

[PhunkySchtuff]

Ummm. No.
That enclosure doesn't do RAID, it's a JBOD enclosure. The peak transfer rate for the mini-SAS interface is 3Gbs (3 Gigabits, not bytes, per second) this is an absolute maximum of 375 MB/sec. The real-world performance of the unit will then depend on the RAID card you're using and will typically be somewhere lower than the peak theoretical performance of the interface. I don't know what drives you're putting in there that can each do 500MB/sec (SSD?) and I don't know what RAID card you propose to use that'll let all eight SSDs run at their peak rate.

The unit I was talking about (http://www.areca.com.tw/products/thunderbolt.htm) on the other hand, with 8 drives in it has a measured real-world performance of 650MB/sec read or write via a single Thunderbolt cable, using RAID 5 that's done in hardware in the enclosure itself.

This 650MB/sec is the actual performance that the BlackMagic Disk Speed Test gave me on a MacBook Pro 13" laptop connected to the RAID with 8x 1TB Western Digital hard drives in it.

Thunderbolt is faster than SAS, SATA and SATA II. Thunderbolt is faster than 2, 4 and 8 Gb/sec Fibre Channel - Thunderbolt is a 10Gbs full-duplex interface, so can transfer 20Gb/sec at it's peak. That's 2.5 Gigabytes per second (1.25 in each direction).

Re: Adoption by Mass Market?

[dbIII]

Doesn't an external SAS connector deliver stuff that fast it's hooked up to something with enough spindles?

Re:Adoption by Mass Market?

[dbIII]

Firewire can do a lot more than two, initially was vastly faster and could do reliable streaming (used to be VERY important with video cameras for one thing, and even USB CDROM burners used to be shit while firewire ones worked due to the reliable streaming). USB won because it was a lot cheaper and usually good enough.

Re:Adoption by Mass Market?

[smash]

Show me a laptop with SCSI, fibre channel or 10GBe please.

Re:Adoption by Mass Market?

[Anonymous Coward]

If your random access use case involves only very small data transfers, perhaps. But if you ever need to transfer a lot of data, like initially copying your files, USB 2 is really slow, topping out around 35 MB/sec if you're lucky. Even FireWire 400 from more than ten years ago will comfortably move data at very close to 50 MB/sec and within spitting distance of the theoretical capacity of the bus. FireWire 800 is twice as fast, both in theory and practice. Simply loading a few hundred GB of data over USB 2 is agonizing. With FireWire 800 it's merely annoying.

FireWire was also nice because at least on Macs it delivered enough power through the bus to comfortably power any 2.5" hard drive, which USB can usually do, but not always, which is why drive makers often bundled those funky pigtail USB cables for their portable drives. Finally, the peer-to-peer nature of FireWire was also very useful on Macs because it allowed most Macs to act as an external hard drive when connected to another computer over FireWire. This was a very compelling feature.

Fortunately USB 3 is a huge improvement. It relieves many of the frustrations of USB 2, outperforms FireWire 800, doesn't suffer the live-plugging issues that have plagued a lot of eSATA on Macs, and due to its very low cost is destined for the same ubiquity enjoyed by USB2. Unfortunately, to the best of my knowledge, USB 3 still doesn't support peer-to-peer transfers, so Macs won't enjoy "target disk mode" over it, nor will it be appropriate for networking or a variety of more rarified use cases that don't necessarily involve a clear master/slave relationship between devices (if USB 3 does support peer-to-peer operation, please comment).

Re:Adoption by Mass Market?

[TheSync]

Show me a laptop with SCSI, fibre channel or 10GBe please.

Well there if you have Thunderbolt, you can do 10GBASE-T:
ThunderLink NT 1102 (10GBASE-T) 10Gb/s Thunderbolt (2-port) to 10GbE (2-Port).

Re:Adoption by Mass Market?

[Anonymous Coward]

Have you ever booted up a second system from your laptop's internal drive using one of its USB ports in "target disk mode"?
It can be very handy for repairs. Firewire and now Thunderbolt allow this.

Re:Adoption by Mass Market?

[tlhIngan]

So, will we see OEM Windows PCs come by default with Thunderbolt ports? Or is this another fantastic, magical, extraordinary Apple Inc. exclusive?

There are laptops coming with Thunderbolt.

Sony's integrated one where the "mobile" mode is a standard Intel 4000 graphics for low power, but then you can dock it (via proprietary USB connector - grr...) which adds a Blu-Ray optical drive AND a decent GPU to the mix. Some company is actually making a PCIe enclosure so you can drive an external card through thunderbolt.

Heck, that's one of Thunderbolt's interesting applications - you can wire up a PCIe video card to it and have powerhouse graphics that suck down the watts, but easily unplug when you don't need it. Essentially, it's a form of hot-pluggable PCIe. And it lacks all the funkiness that USB adapters typically entail.

A thunderbolt serial port, while overkill, will present itself to your laptop as a NATIVE serial port - no messing with icky USB serial adapters that are iffy - this works just like a built-in serial port because it is using the standard busses your PC expects. As far as anyone is concerned, it hooks straight to the PCIe bus, and does normal PCIe things, and other than some minor hardware bridging, acts like it's plugged into an internal PCIe bus.

Re: Adoption by Mass Market?

[PhunkySchtuff]

It can do these kinds of transfer rates, however SAS enclosures (with built-in RAID controllers) tend to be more expensive and then you also need a SAS interface card, whereas Thunderbolt is now being built into motherboards.

This is starting to get to the upper limits of what SAS can do. Only Fibre Channel and Thunderbolt will do these kinds of rates with room to grow.

Re: Adoption by Mass Market?

[Anonymous Coward]

* Slaps forehead *

Re:Adoption by Mass Market?

[OhANameWhatName]

Don't forget the ability to chain drives, that was pure awesomeness.

Re: Adoption by Mass Market?

[Mad Merlin]

The peak transfer rate for the mini-SAS interface is 3Gbs (3 Gigabits, not bytes, per second) this is an absolute maximum of 375 MB/sec.

I'm sorry, but you're wrong. Have a look at this review, for example.

Each mini-SAS cable provides four lanes of SAS (3 Gbit/s), SAS2 (6 Gbit/s), or SAS3 (12 Gbit/s), depending on the HBA in use. That equates to 12 Gbit/s, 24 Gbit/s or 48 Gbit/s per cable. Also, with SAS2 being out since 2009, it's pretty hard to even find a SAS1 card anymore.

Re:Adoption by Mass Market?

[Chrontius]

Firewire was great because USB2 had a problem of shared bandwidth, and the sliver of my total 480 MB/sec allotted to my backup drive and iPod made Firewire's 400 MB/sec of intelligently allotted bandwidth a lot more compelling.

On top of that, by the time I got the iPod, the Pentium 4 was showing its age; anything that made the whole experience more responsive was appreciated.

Plus, I was quite an impatient little tyke at the time. I haven't gotten all that much better, but the technology sure has.

Re:Adoption by Mass Market?

[Anonymous Coward]

Show me a laptop with SCSI, fibre channel or 10GBe please.

Perhaps you're unaware that USB uses SCSI for its transport channel.

Perhaps /dev/sda1 would give you a hint.

Re: Adoption by Mass Market?

[bored]

The peak transfer rate for the mini-SAS interface is 3Gbs (3 Gigabits, not bytes, per second) this is an absolute maximum of 375 MB/sec

Uh, that Sans Digital has a SFF-8088 connector, which is a x4 SAS link. AKA SAS is like PCIe, you can gang lanes, and that is exactly what this is doing. So its actually 12Gb/sec not 3. The hard drives are individually 3Gbit, but the backplane is a SAS expander.

Connect it to a RAID card, and you should be able to pull pretty nice numbers (they claim close to 800MB/sec). That said, Sans digital tends to be a little low end. Modern SAS JBODs are 6Gbit SAS, many come with a pair of x4 links (48Gbit for those counting). So, its completely possible to pull >2GBytes/sec out of a low end SAS/JBOD connected to a ~$250 RAID controller. Double that (~4GB/sec as the GP said) if you want to spend some money and buy a real RAID.

If your really cost sensitive, you can get a good raid controller with 4-8 internal connectors and one of the dozens of (see supermicro for example) cases with internal SAS/Sata expanders for a few hundred more. If you pick the right motherboard it might even be integrated. These machines can easily break 1.5GB/sec without even trying.

Thunderbolt is faster than SAS, SATA and SATA II. Thunderbolt is faster than 2, 4 and 8 Gb/sec Fibre Channel

You of course realize that FC is currently 16Gbit full duplex, and FCoE is available at 40Gbit, and that its a switched topology with trunking and long distance interconnect capabilities? Plus there are "proprietary" up-link extensions like the qlogic XPAC format that runs at 64Gbit full duplex? Here is a switch that supports all of the above. (http://www.qlogic.com/Products/Switches/Pages/ConvergedSwitches.aspx).

Furthermore, FC cards regularly come in 2 and 4 port models. And the boards support NPIV, PCI SR-IOV, load balancing, and other vitalization technologies that allow you to mix and match the physical and virtual port layouts to split/combine bandwidth to a single server as necessary.

And if that isn't enough, there is still infiniband. See mellanox's latest offering.

The bottom line, is that thunderbolt is a nice cable for laptops, but it in no way competes with modern interconnects available on servers and high end workstations. Especially for storage and networking, if for no other reason than the existing systems are switched and allow multiple adapters per machine. I have choked the memory/bus system on high end server class machines by zoning a large number of FC disks to a single node with multiple adapters. The memory/PCIe bandwidth on a quad socket machine gets maxed out before the FC port count gets exhausted.

Given the prevalence of low end SAS/SATA port expanders and 4-32 channel RAID cards it doesn't really look particularly inexpensive either.

Re:Adoption by Mass Market?

[Anonymous Coward]

Smash dumb. Smash read post all wrong. Why Smash can no read simple post?

Thunderbolt is great for laptop users who can't have that stuff on-board. Most desktop workstation users don't need it.

Re: Adoption by Mass Market?

[Anonymous Coward]

It's even more pathetic because jedidiah and h4rr4r are fervent religious zealots that can't actually write code. They're basically everything that's wrong with GNU.

Re:Adoption by Mass Market?

[Anonymous Coward]

For SCSI, all the Powerbooks up to the first Powerbook G3 (with maybe a couple of exceptions like Duo's) had SCSI built in... I think my boss's Micron (Pentium MMX 150MHz) had a SCSI port,too. I don't know of any modern laptops with SCSI, though. Too bad it didn't catch on.

Re:Adoption by Mass Market?

[Anonymous Coward]

USB requires more CPU during transfers. This is also a big no no unless you don't care.

Re:Adoption by Mass Market?

[jo_ham]

Have you ever seen a Thunderbolt port?

I guess not, based on your post. Still, don't let the facts get in the way of a good bash.

Re:Adoption by Mass Market?

[bloodhawk]

And that is just it. Good enough is all most consumers want, especially when their is a price differentiator. At the moment the average consumer doesn't have a desperate need for thunderbolt, that isn't to say it isn't better, just this really is still borderline as to wether it falls through the cracks to become another firewire or niche market. Consumers simply don't demand the absolute best, they demand good enough for the money.

Re:Adoption by Mass Market?

[dbIII]

I'm not so sure this time. People buying 4TB or soon larger external drives are not going to be so happy waiting for them to fill up with movies or whatever at USB speeds (especially since there's not a lot of USB3 in laptops), so may fork out for the more expensive technology. That may establish it as something other than a thing to plug overpriced Apple screens into.

Re:Adoption by Mass Market?

[Anonymous Coward]

Most people don't. The CPU usage of USB is still very low.

Re:Adoption by Mass Market?

[TheRaven64]

Back in 2003, I had a PowerBook that came with FireWire 400 and FireWire 800. I had two LaCie BigDisks, which had USB, FireWire 400 and two FireWire 800 ports. I could connect one to the other and then daisy chain them to the FW800 port. USB 2 was the bottleneck when using a single drive. With FW400, the bottleneck became the drive speed. With FW800, I could have both drives running at their full speed. This was amazing for video editing on the laptop. I had the source DV files on one disk and the scratch disk on the other, and compositing passes became CPU-bound, rather than I/O bound (as they were to a painful degree on the internal disk drive). Back then, half decent video cameras were also FW400-only, and with an adaptor cable I was able to plug the camera into the end of the chain (I only did this once to see if it worked, because I had another FW400 port for it in normal use) and it worked perfectly, streaming the data to a disk on the same chain. On that machine, I used the FireWire ports a lot more than the USB ports - I initially used them for a keyboard and mouse, but later replaced those with Bluetooth.

It wasn't a separate technology required for hard drives, it was a separate technology required for things that needed either high bandwidth or isochronous transfer. Unfortunately, that basically meant hard disks and video cameras. It failed because USB 2 was only slightly worse than FW400, and you needed USB (and so got USB2 effectively for free) for keyboards and other cheap devices. A FireWire mouse or keyboard would have needed the $1 host controller chip, whereas a USB one could use much simpler hardware, so would always be cheaper. You might have got less lag with a FireWire input device, but I doubt anyone human would have noticed the difference and been willing to pay the price (although you might have been able to sell them to gamers). USB was in the Intel chipsets, FireWire was an extra IC on the motherboard, so it added cost. It's hard to make people pay extra to go from good enough to better. The proliferation of USB2 ports and fast CPUs made this even worse: the issues with USB2 largely go away when you have one port per device and enough cores that you can effectively designate one to running the USB devices and still not notice.

Re:Adoption by Mass Market?

[Leejjon]

Well my Sony Vaio has an usb compatible thunderbolt port, which only can be used to connect the external video card, and as a regular usb 3.0 port. I think Sony is paying Apple for a license to use the technology. Apple could easily make the thunderbolt port work as an usb port too, but they just don't because they want to sell extra cables.

Re:Adoption by Mass Market?

[EvilSS]

I think you mean Asus HAD a motherboard. The p8z77-v premium has been discontinued.

Re:Adoption by Mass Market?

[wagnerrp]

USB2 was considerably worse. Even rated at 480Mbps, FW400 would still run circles around it. The trouble was that as mentioned elsewhere, even the lackluster performance of USB2 was "good enough" for the mass market. Where USB was CPU driven, and thus dirt cheap to implement, Firewire required real hardware and DMA, along with a significant licensing cost, and the couple dollars it added to every piece of equipment that implemented it sunk the format.

Re:Adoption by Mass Market?

[kevingolding2001]

No. Yes.
Thunderbolt is the next Firewire.

Re:Adoption by Mass Market?

[otuz]

What would they pay Apple for? It's Intel tech and most PC buyers/manufacturers are just such cheapasses they don't want to pay extra for things like that.

Re:Adoption by Mass Market?

[bloodhawk]

The limiting factor with external drives is not the transfer rate of the connection, it is the speeds of the drive itself, you could have a 100TB/s connector, it still isn't going to run at any significantly faster rates. Current 7200 speed 4TB drives max out under 200MB/s

Re: Adoption by Mass Market?

[Anonymous Coward]

man I am not sure what is sader, your complete lack of knowledge of SAS and fibre channel bandwidth or the fact you got marked as informative. For fuks sake go read up on stuff before you blurt Bullshit into your keyboard.

Re: Adoption by Mass Market?

[fnj]

Err, it says it has TWO SFF-8088s.

Re:Adoption by Mass Market?

[dbIII]

None of the USB stuff I've seen had gotten close to SATA speeds yet. Either I've got crappy USB3 hardware on half a dozen fairly new machines, crappy USB3 drivers on a couple of different operating systems or maybe USB3 in general hasn't delivered on that promise yet. Thunderbolt has a pretty low bar to jump over as far as I can see.

Re:Adoption by Mass Market?

[Kjella]

Firewire can do a lot more than two, initially was vastly faster and could do reliable streaming (used to be VERY important with video cameras for one thing) (...) USB won because it was a lot cheaper and usually good enough.

Firewire was a must have for (H)DV cameras that recorded to tape, where you absolutely had to catch the frame as it was played off. Even so it was a total pain to get a perfect capture, I had to kill all background applications and couldn't do anything else with my computer or I'd lose the odd frame here and there. And it'd run at the same speed as recording, one hour to record meant one hour to transfer. The moment reasonably priced flash/HDD based cameras became available recording to tape was dead and Firewire with it, because the bandwidth on USB was fine once it could do flawless transfers.

Re:Adoption by Mass Market?

[nonicknameavailable]

Thunderbolt is the new firewire 800

Re: Adoption by Mass Market?

[Anonymous Coward]

really? Do you think 'it is value and appeal is pretty dubious' is correct? Seems to me that a possessive 'it' is correct, rather than what was posted. Talk about a knee jerk reaction.

Re: Adoption by Mass Market?

[jedidiah]

> What exactly are the poor engineering choices

Systems that NEED an external card cage to make up for things that are missing in the original product.

"Worry not if that low profile machine or laptop has a GPU mocked by everyone. You can buy a big ugly expensive box to attach to your pretty little Apple."

Re: Adoption by Mass Market?

[jedidiah]

> note: jedidiah is a gnu/hippie who's angry that Apple took-over the *nix desktop market.

MacOS is it's own proprietary OS and always has been. To what degree it has Unix embedded in it (for corporate cost savings) has about as much relevance to the typical user as the Linux kernel running in a Tivo or Sony BluRay player.

It must really suck to be a MacOS fanboy. They make hardware that's either lame and limiting or really expensive. That has to be the worst part about wanting to run MacOS: needing to buy Apple hardware.

Macs suck as PCs. It doesn't matter if you've deluded yourself into believing they're Unix workstations.

That's the fun part about all of this. If you've bought a serious Mac you are locked out of all of this shiny newness. You can't get an upgrade card for it either.

Re: Adoption by Mass Market?

[jedidiah]

650MB/s? That's all? I can get nearly that much just by putting drives into a PC. I get nearly that much from the drives in my PC.

You're out of touch with what PCs can do if you think those numbers are going to blow the rest of us away.

Doing that over a distance would be more interesting.

Re: Adoption by Mass Market?

[jedidiah]

I get more than 375MB/sec (total) from cheap SATA ports bundled on a cheap consumer PC motherboard. The pro grade stuff has to be better than that.

A cheap consumer spinny drive can manage 200MB/s by itself.

It's a shame that 10G ethernet is not cheap and commonplace.

Re: Adoption by Mass Market?

[drinkypoo]

A cheap consumer spinny drive can manage 200MB/s by itself.

For what, sequential reads on a drive without remapped sectors? In the real world it doesn't do that, or at least, you can't expect it to.

It's a shame that 10G ethernet is not cheap and commonplace.

1G ethernet has only recently become cheap and commonplace, but it's good enough for most uses for most users.

Re: Adoption by Mass Market?

[coinreturn]

really? Do you think 'it is value and appeal is pretty dubious' is correct? Seems to me that a possessive 'it' is correct, rather than what was posted. Talk about a knee jerk reaction.

You forgot to call him a Grammar-Nazi Nazi.

Re:Adoption by Mass Market?

[drinkypoo]

Maybe the next generation of graphics cards will be external, to isolate it from CPU and PSU heat,

...and then this interface still won't be fast enough to talk to it satisfactorily, and is useless for that. A graphics card that powerful will need more bandwidth than you can get over this link. It's not x16.

future-proofing, followed by an incremental upgrade, is often cheaper than buying two or three mid- or low-end systems over the course of the computer's life.

That's true. That's why I didn't buy an Apple machine. I bought pieces and built. I got a tri-core processor because it was less than half the cost of the six-core, then I incrementally upgraded later. I also got the system without USB3 because at the time it was an expensive motherboard option, and added USB3 later for fifteen bucks. Since you bought a PC, you could have got a much cheaper motherboard, and then upgraded the motherboard to one which had Thunderbolt later, if you actually ever used it. I've known several people who went out of their way to get a PC motherboard with firewire onboard and then literally never used it once. They could have added it later with a fifteen dollar peripheral card, or added it never and saved twenty or thirty bucks.

Only Apple or mobile device users have to buy all their features up front. The rest of us can upgrade.

Re:Adoption by Mass Market?

[TheSkepticalOptimist]

Why, most of the exploits these days are on OS X anyways.

Re:Adoption by Mass Market?

[drinkypoo]

The trouble was that as mentioned elsewhere, even the lackluster performance of USB2 was "good enough" for the mass market.

How is that trouble? That's simply a testament to the viability of USB2. I've had far more problems with firewire than with USB2 (flaky drivers, flaky chips, flaky cards) so even though it is much better in the performance department I gave up on it even before the industry did. Notably, daisy chaining doesn't work on cheap hardware and/or with cheap cables. If I could afford expensive hardware, I could afford a firewire hub. Firewire may be faster than USB, but frankly I think that overall USB is "better" than firewire. And I have gone out of my way to make sure that all my PCs have had it for years. Thing is, now I have USB3 disks and I don't need it any more. And frankly, even over USB2 it's not unusual to see 40MB/sec to a HDD. In terms of real-world performance on a segmented disk that's fine for most purposes for most people. If you need more bandwidth there's eSATA. My PC has external USB2, USB3, Fw400, eSATA, and SATA ports. (The latter not being hotswap, but available on my memory card reader front panel thingamagig.) I haven't even used the firewire on it. Not worth the hassle.

Re:Adoption by Mass Market?

[drinkypoo]

Show me a laptop with SCSI, fibre channel or 10GBe please.

I hope you mean a modern laptop, because laptops used to have SCSI all the time. And, it was good. The fastest SCSI I've ever seen on a laptop was, I think, fast-narrow, but I'd bet someone made something with ultra-narrow. I had a fast-narrow PCMCIA card for a while, Adaptec-made and zip-branded, but you couldn't boot from it.

Re:Adoption by Mass Market?

[wagnerrp]

My experience has always been the opposite. The handful of firewire components I've had always "just worked", while any heavy USB usage was liable to cause a kernel panic.

Re:Adoption by Mass Market?

[Anonymous Coward]

Nice, your comment is completely semantically null - well done.

Re: Adoption by Mass Market?

[Anonymous Coward]

Both wrong, SAS does 6Gbps (soon 12Gbps) with ease, over that connector. How do I know this? I own the enclosure, and it performs quite nicely connected to my LSI 9280-4i4e.

PLUS - Thunderbolt is useless for storage on anything other than desktops - it's not a storage interface, it's PCIe. So what, you say? Well, each "endpoint" is really a controller with disks attached to it. Unless they are single disks, which would kind of be a travesty anyways, you can't run a real filesystem over multiple drives (ZFS does not tolerate lies), and you can't use the disks in conjunction with a traditional RAID controller either.

Everything else that Thunderbolt does is either an egregious waste of bandwidth (USB / FireWire / Ethernet? USB 3 is more than enough), or not enough bandwidth (Video card? PCIe x2? Don't make me laugh), or redundant (DisplayPort works just fine without Thunderbolt, thank you). It's only Apple's fetish for shiny + the extra dollars from the fact that only THEY make the magic cables that keeps this standard going.

In addition, four Vector SSDs connected to the same controller stomps the crap out of that 50-drive Thunderbolt array that I believe Ars did a story on a few weeks ago. It's true, I get sustained bandwidth of about 1.7GBps down and 1.8GBps up, and don't need to warp a perfectly good folding table to do it.

Re:Adoption by Mass Market?

[Rockoon]

'cept how are they going to acquire these movies at greater than USB speeds, in order to need greater than USB speeds?

For example, BLURAY drives stream 54Mb/s at 1x speed, and the fastest common drives are 12x, leading to 648Mb/s which is 81 MB/s.

USB 3.0 has a theoretical maximum transfer rate of 4.8Gb/s which is 600 MB/s, over 7 times faster than 12x BLURAY drives and over 42 times faster than the commonly needed 2x speed for realtime playback of HD content (42 minutes of HD BLURAY video could transfer over USB 3.0 in 1 minute)

So where are they magically going to be getting these movies that frustrate them about how slow their external bus is? In short, you don't seem to know what you are talking about but instead are throwing around crap trying to justify a common-mans decision to go with an overly expensive external bus. It doesnt wash because the common man has no use for the added speed. The fact is that the common man doesnt even have a hard drive rated at faster than USB 3.0 speeds, and the drives that do rate faster are so expensive that the common man would never ever buy one.

Re:Adoption by Mass Market?

[serviscope_minor]

A thunderbolt serial port, while overkill, will present itself to your laptop as a NATIVE serial port - no messing with icky USB serial adapters that are iffy - this works just like a built-in serial port because it is using the standard busses your PC expects.

You can buy whatever the modern PCIe equivalent of PCMCIA is serial port cards. They work very well. Much better than those nasty USB ones.

Re: Adoption by Mass Market?

[Anonymous Coward]

Oh, yah I missed that (dual 8088s). I looked closer at the specs and its not clear to me if its just a port mirror or they can be configured into a single link. Given my experience with Sans digital (Yah, I own a 5 bay at home) they aren't always stellar performers so it wouldn't surprise me if the configuration is stupid in some manner. They claim 750MB/sec max, which may just be because they didn't plug in any SSDs, or it could be some internal limitation. The 750MB/sec sounds like what you could expect from a 8 disk (spinning media) RAID5. So it could be either way.

Re:Adoption by Mass Market?

[Rockoon]

Turns into a circular argument that is easily defeated...

A: "There is no demand for SCSI, fibre channel, or 10GBe in laptops, so there wont be demand for Thunderbolt."
B: "But there are adapters that convert Thunderbolt to those"

A: "That doesnt create demand where there wasn't any before."

Re:Adoption by Mass Market?

[smash]

Exactly my point. Portables with "SCSI (i.e., SAS), fibre channel or 10GbE dedicated ports do not exist. Hence, having a general purpose 10 gigabit external extension of the PCIe bus is a nice thing to have.

Re:Adoption by Mass Market?

[smash]

Perhaps you're unaware that USB is a CPU driven bus that sucks for performance, never mind what software stack is running it.

Re: Adoption by Mass Market?

[Anonymous Coward]

All hail JOBS!

Re:Adoption by Mass Market?

[smash]

Yeah, i meant SAS. i.e., serial scsi. Point being, no other bus gives you an external PCIe capability like thunderbolt does - which enables you to run pretty much any sort of device you could plug into a PCIe slot.

Re: Adoption by Mass Market?

[Khyber]

"This is starting to get to the upper limits of what SAS can do. Only Fibre Channel and Thunderbolt will do these kinds of rates with room to grow."

And like your prior comment, this one's dead wrong too.

Wireless 4K *UNCOMPRESSED* HDMI @120Hz with ARC, 3D, networking, Deep Color, and remote storage data pipelines.

Thunderbolt can't even fucking compare.

Re:Adoption by Mass Market?

[Khyber]

Your captcha was calling you the retard.

USB3 took forever to get here, we had Firewire 3200 for fucking AGES.

Go figure it's an anonymous coward that's wrong again, as usual.

Re: Adoption by Mass Market?

[UnknowingFool]

[sarcasm]Yes because all laptop manufacturers include TB RAID arrays with the laptops except for Apple.[/sarcasm]

Thunderbolt is Intel's answer to this problem as well as the lack of a universal laptop connector and the lack of external data connection. As far as I know, most docks don't have a high bandwith data external connection like eSATA. It competes against eSATA, SAS, and in some use cases USB3. Whether it will succeed is a different story. Apple will use it as it solves their problems just like they used FireWire. If you are short on memory, at the time Apple started using it for the iPod, USB 2 hadn't been ratified yet. You either had serial or USB1.1 (max rate of 12 Mbs) as the connection. When it USB2 was more widespread, Apple switched to it. Another reason was that Apple could replace the FireWire chip with a video one. [sarcasm]But there's Apple making poor engineering choices again.[/sarcasm]

Re: Adoption by Mass Market?

[UnknowingFool]

That's on the laptop manufacturer. They chose to make it like that. The Thunderbolt spec from Intel does not require such limitations. Intel owns TB not Apple. Apple was just the first to use it and is advertising the most about this feature.

Re:Adoption by Mass Market?

[Anonymous Coward]

Thanks, Apple.

For the record, you're an ignorant buffoon.

Re: Adoption by Mass Market?

[Anonymous Coward]

"Poor engineering choices"

LOL, 4 digits and still a dickhead.

Re: Adoption by Mass Market?

[tehcyder]

Nope, It's == it is, its == belonging to it. If you're going to try to be a Grammar Nazi, at least be right.

The irony here is too strong for the human mind to bear. I feel like some poor space miner accidentally finding themselves stepping in a black hole of stupidity.

Re: Adoption by Mass Market?

[Anonymous Coward]

What transfers at 2.5GB/s (even if it could attain it)? Even the best internal SSDs transfer at ~300-600MB/sec (theoretical peak), and most transfers are one way (or at least, large transfer in one direction).

Besides, you conveniently left out SATA3 which has been out for a while now. Drop out external drives, what else needs that kind of bandwidth? I can't think of any.

A 4K monitor you say? Why would I drive the monitor out of the bandwidth when I could just have a remote video card with better performance? Most computers these days also come with at least one secondary monitor port anyway...

Re:Adoption by Mass Market?

[Anonymous Coward]

How often does one "wait to fill" it up? Never really.

Even with bittorrent, downloading straight to an external will never even touch max throughput on a USB2. Hell, while I was downloading a few things from BT, I was watching a 720p video off of my old computer (dual Xeon 2.4Ghz, with a 1TB external).

Even if I'm copying something rather large to the external that will take a few minutes, I just start the transfer and then go do something else. Besides "most people" will not be filling them up with movies -- they stream them all these days. Just the tech nerds will download and possibly protest about DRM.

Re:Adoption by Mass Market?

[Anonymous Coward]

Too true. Seven years ago I had a firewire hard drive and dvd drive. While daisy chained I was able to rip cds from the drive, encode them, and write them to the hard drive at a speed which I sadly have never been able to match to this day. I don't need that kind of performance day-to-day though so I'm using usb2 for everything due to the price difference and availability.

Re:Adoption by Mass Market?

[Anonymous Coward]

Who cares? I just get a $10-15 5-10 port USB hub. Problem solved.

Sure, it's nice to daisy chain, but really? That's the selling point?

Captcha: Maleness

Re:Adoption by Mass Market?

[Anonymous Coward]

I have a GPS logger that presents itself to the desktop as a COM port. I've plugged them into numerous Windows based laptops and desktops over the course of the past 4-5. Never once have I had a problem with drivers.

But hey, if you buy "icky TB serial adapters that are iffy"... You might even have worse trouble BECAUSE it's presenting itself on the PCIe internal bus!

Also, "icky"? What are you, 5?

Re:Adoption by Mass Market?

[Joce640k]

Have you ever seen a Thunderbolt port?

I guess not, based on your post. Still, don't let the facts get in the way of a good bash.

You mean this?

Yeah, that's way different than all those other connectors. Not.

Re:Adoption by Mass Market?

[jo_ham]

Well my Sony Vaio has an usb compatible thunderbolt port, which only can be used to connect the external video card, and as a regular usb 3.0 port. I think Sony is paying Apple for a license to use the technology. Apple could easily make the thunderbolt port work as an usb port too, but they just don't because they want to sell extra cables.

Again, the misinformation around thunderbolt is hilarious.

Apple DOES NOT OWN the Thunderbolt technology. It belongs to Intel. There is no requirement to make it fit into a USB port, or to only work that way. The official spec calls for the use of the mini displayport connector, which Apple owns the licence to, which it has granted royalty free licencing to anyone who wants to use it. The MDP connector looks nothing like USB.

Sony is not paying Apple anything to use the technology.

Re:Adoption by Mass Market?

[jo_ham]

Have you ever seen a Thunderbolt port?

I guess not, based on your post. Still, don't let the facts get in the way of a good bash.

You mean this?

Yeah, that's way different than all those other connectors. Not.

Well done, you have noticed that they very deliberately chose to use the MDP connector.

Not sure if this is sarcasm or genuine buffoonery...

If it is a troll, well played sir.

On the other hand, if you believe the MDP port is "almost indistinguishable" (direct quote) from microUSB and/or microHDMI then I think you should probably box up your computer and return it to the store. It will be better for everyone in the long run.

Re: Adoption by Mass Market?

[Anonymous Coward]

PLUS - Thunderbolt is useless for storage on anything other than desktops - it's not a storage interface, it's PCIe. So what, you say? Well, each "endpoint" is really a controller with disks attached to it. Unless they are single disks, which would kind of be a travesty anyways, you can't run a real filesystem over multiple drives (ZFS does not tolerate lies),

What in the fuck kind of drugs are you on? Typical Thunderbolt disk enclosures are TB-to-PCIe bridge chips plus an off-the-shelf PCIe SATA or RAID controller. Where is the "lie" which ZFS will hate? Not even the SATA or RAID card drivers can tell whether the SATA/RAID chip is sitting on the other side of a Thunderbolt link! Drivers do need to support PCIe hotplug if they're going to be robust in the face of the user unplugging the Thunderbolt cable, but that's not even a Thunderbolt-specific capability. (PCIe hotplug has been around since the dawn of PCIe.)

and you can't use the disks in conjunction with a traditional RAID controller either.

Large Thunderbolt disk enclosures literally are a "traditional RAID controller", moron. Here's an easy example, one of the first Thunderbolt devices to hit the market:

http://www.anandtech.com/show/4489/promise-pegasus-r6-mac-thunderbolt-review/2

It's literally a PMC-Sierra PCIe SAS RAID controller and backplane stuffed inna box with a Thunderbolt bridge.

Everything else that Thunderbolt does is either an egregious waste of bandwidth (USB / FireWire / Ethernet? USB 3 is more than enough),

You are stupid. You're doing the moral equivalent of whining that PCIe itself is a waste of bandwidth when you plug in a gigabit NIC PCIe card because even 1 lane of PCIe 1.x has a lot more BW. Well duh, moron, the idea behind generic high performance system interfaces is to provide a common attachment point for a variety of things. Of course it's going to be faster than some of them! And it's especially important to be lots faster for scenarios when many devices might be sharing the same link, as is the case in a Thunderbolt daisy chain.

And no, USB3 isn't enough. USB3 is fine for single HDDs, more than fine even. But if you want to attach another NIC, it's kinda far from ideal. Unlike PCIe (or PCIe tunneled over Thunderbolt), USB is piss poor at general purpose DMA, low latency interrupts, and so forth. USB wasn't designed for that. It can be done, as the hojillions of USB2 100Mb ethernet pods testify, but it's never as good as a PCIe NIC. A Thunderbolt NIC is, because it literally is a PCIe NIC.

or not enough bandwidth (Video card? PCIe x2? Don't make me laugh), or redundant (DisplayPort works just fine without Thunderbolt, thank you). It's only Apple's fetish for shiny + the extra dollars from the fact that only THEY make the magic cables that keeps this standard going.

Once again you are stupid. There are other sources for cables than Apple. And dumbass hater cant about "SHINY FETISH FANBOY SHEEPLE" is just so ignorant -- it's pretty clear that a large amount of Apple's motivation for adopting the interface was to make it possible for a large chunk of their "pro" userbase to do things with Macbooks which previously required towers. For example, there are plenty of pro audio/video capture/compression/etc. PCIe cards which got put on sale as external peripherals inna Thunderbolt box soon after Apple started shipping portables with Thunderbolt ports. Video pros don't want to be anchored to a 50 pound tower if they don't have to be.

But hey, it's invisible to you because you don't know about anything outside datacenters, so it doesn't exist, amirite?

In addition, four Vector SSDs connected to the same controller stomps the crap out of that 50-drive Thunderbolt array that I believe Ars did a story on a few weeks ago. It's true, I get sustained bandwidth of about 1.7GBps down and 1.8GBps up, and d

Re: Adoption by Mass Market?

[Anonymous Coward]

And like your prior comment, this one's dead wrong too.

Wireless 4K *UNCOMPRESSED* HDMI @120Hz with ARC, 3D, networking, Deep Color, and remote storage data pipelines.

Thunderbolt can't even fucking compare.

I am not PhunkySchtuff but you, sir, are a fucking idiot. Every feature you mention is enabled by exactly one thing, bandwidth. And Thunderbolt has more of it than HDMI. A lot more of it.

Today's Thunderbolt: Two full-duplex 10 Gbps links in one cable. Accounting for the 64b66b line coding overhead, that's a grand total of 19.4 Gbps in, 19.4 Gbps out. One of the full-duplex 10G links is typically used for encapsulating DisplayPort video packets, the other for encapsulating PCIe packets (yes, both DP and PCIe are packet busses).

Today's HDMI (1.4): Three half-duplex 3.4 Gbps links, dedicated to outbound video (one per primary color). Line coding is the less-efficient 8b10b, so this translates to a total of 8.16 Gbps outbound. To support "ARC" and networking, there's a low performance full duplex side channel of some sort. Apparently it supports ~37 Mbps audio and 100 Mbps ethernet, so call it 200 Mbps total.

So, it's not even close. A single thunderbolt port can transport more bits per second of video data than HDMI 1.4 while simultaneously connecting to a 5 Gbps disk array, a GbE Ethernet NIC, and as many high resolution audio ADCs or DACs as you care. With lots of spare bandwidth left over. HDMI has shitty 100 megabit networking and "audio return" clumsily retrofitted in because it was originally a one-way interface. (Literally the only path for data to flow inwards in the original HDMI was the EDID monitor type detection.)

Tomorrow's HDMI (2.0) raises the raw outbound rate to 18 Gbps, while Thunderbolt goes up to a raw 20 Gbps per link (still 2 links). So it's only going to get skewed further in favor of Thunderbolt.

Also: HDMI 1.4 doesn't support 4K at 120 Hz. Try 24 Hz instead. Even the next generation HDMI 2.0 is only going to do 60 Hz 4K video. (or 30 Hz if you're doing "3D".)

Re:Adoption by Mass Market?

[dbIII]

Discs.

Re:Adoption by Mass Market?

[dbIII]

To explain the obvious - I was writing about a copy from PC to external drive. With slow USB2 drives it can take a full working day to put much less than a TB on an external disk, and so far USB3 for whatever reasons doesn't seem to be much faster. If thunderbolt is getting it right early on then there's a place for it.

Re: Adoption by Mass Market?

[Khyber]

Thunderbolt is STUCK AT 10Gbits per channel (with a maximum of 20 channels) while HDMI can go much higher than that, and 48-bit deep color support pretty much ensures you need more bandwidth than Thunderbolt could ever provide on a single channel, (not enough bandwidth to support trillions of colors @ 4K resolution) while HDMI has no issues with it.

Re: Adoption by Mass Market?

[Anonymous Coward]

Thunderbolt is STUCK AT 10Gbits per channel (with a maximum of 20 channels) while HDMI can go much higher than that,

You are such a low effort troll and liar. Hint: you're in a discussion attached to a Slashdot article about the announcement of a next generation 20 Gbps per channel Thunderbolt. Another hint: There is nothing magic about HDMI which permits it to scale endlessly. In fact,

and 48-bit deep color support pretty much ensures you need more bandwidth than Thunderbolt could ever provide on a single channel, (not enough bandwidth to support trillions of colors @ 4K resolution) while HDMI has no issues with it.

Bullshit. HDMI 1.4 supports 8.16 Gbps outbound. 4K @ 48-bit color and 24 Hz refresh is 3840 * 2160 * 48 * 24 = 9.56 Gbps.

Re: Adoption by Mass Market?

[Khyber]

>24 FPS

>120Hz Plasma 4K can do

And there's where you're pretty stupid.

Re:Adoption by Mass Market?

[Anonymous Coward]

lol, even though I admitted Firewire's superiority, I still get a angry firewire zealot replying.

Look: if you actually cared about raw transfer speeds, you wouldn't be using FW either. Joke argument unless you're living in the mid-2000s.

USB2 is fine for consumer storage and backups. And if you're on Windows, Firewire is just another glitchy problem (at least in my experience). USB3 is here now and super-advanced-firewire-120000 isn't. End of story.

captcha: retard

You thought I was an angry zealot? I apologize. I'm not. I was pointing out that there are advantages to FireWire that might manifest themselves even in simple scenarios, and that USB 3 is the answer to both in most respects. I admit I erred by not pointing out that the ubiquity and low cost of USB 2 spoke in its favor. At least using a Mac, I'd prefer FireWire over a USB 2 drive, especially if I didn't have to pay the extra $50 they tend to cost these days. There are certainly faster interfaces, but most of them aren't built in nor even very realistic choices for most Macs (for reasons I alluded to, as well as Apple's obstinacy in not making many Macs with expansion ports, at least prior to Thunderbolt which itself isn't much of an improvement for home users or anyone on a budget). A home user on a Mac gets to choose between USB and FireWire before the price more than doubles, and depending on the Mac may not enjoy any other choices. Users not running Macs also get to consider eSATA. Anyone with an unlimited budget can have whatever they want.

Since this is a technical site where there was a comparison being made between USB and FireWire, I felt it was appropriate to go in to more detail about the capabilities of FireWire. Not everyone knows about all these features, but they can be useful even in low-end scenarios.

Let's summarize:
USB 2 - Inexpensive, ubiquitous, highly compatible, not quite quick, maxes out at relatively little bus power, and doesn't support peer-to-peer operation (no "target disk mode")

FireWire - Significantly more expensive, very well supported on Macs, mediocre to spotty support on Windows, not widely deployed, somewhat to much faster than USB 2 depending on version, peer-to-peer operation supports networking and Apple's "target disk mode," bus power varies depending on vendor implementation from none to a lot (on Macs, a lot) at a voltage decided by the vendor

USB 3 - Very fast, destined for ubiquity, more bus power (standardized), all the advantages of USB 2, still not peer-to-peer (to my knowledge). USB 3 is default "right choice" for someone who has USB 3 ports and doesn't have more specific needs.

eSATA, iSCSI, ATA over Ethernet, SAS, Fibre Channel, Thunderbolt, PCIe, etc: Too much to compare here

Oh, bitches!

[Anonymous Coward]

Yes!

What could I connect this to?

[chromaexcursion]

Apple has pissed off all the other CE manufacturers. There will be nothing to plug the other end into.

Without general support great features are worthless. Apple is repeating Sony's mistake with betamax. They won't share, thus it will fail.
Great technology without support is worthless.

Re: What could I connect this to?

[Anonymous Coward]

But isn't Thunderbolt an Intel technology?

Re:What could I connect this to?

[Anonymous Coward]

Apple has pissed off all the other CE manufacturers. There will be nothing to plug the other end into.

Without general support great features are worthless. Apple is repeating Sony's mistake with betamax. They won't share, thus it will fail.

Great technology without support is worthless.

iTV, coming to a stage near you!

Re:What could I connect this to?

[Radagast]

That's simply false. There's a large amount of Thunderbolt accessories, including video gear, PCIe expansion chassis (very useful for laptops), and docks. Sonnet just announced this Thunderbolt dock, which seems to be a pretty great deal for laptops.

Re: What could I connect this to?

[UnknowingFool]

Repeat after me: Thunderbolt is an Intel technology. It is the copper version of LightPeak. Apple was the first to have it in their products because they worked with Intel on a number of things like the mini-Display Port connector. Thunderbolt == Intel.

Re:What could I connect this to?

[jedidiah]

...for differing values of large.

I am sure that you would also claim that Firewire rules the world too.

With all of the costs and extra gear involved. You might as well just have another PC. The real problem here isn't that Apple laptops are lame but that there isn't a seamless experience between different devices on an Apple network.

The Cloud concept there doesn't quite live up to the hype.

Re:What could I connect this to?

[Idimmu Xul]

I want one, but srsly, the cost of that thing is retarded. $400 for the base option, which is essentially an extravagant usb3.0 hub and dvd rw with some bells on it.

Re: What could I connect this to?

[chromaexcursion]

except only apple is the only real supporter.
Intel made it for apple. Intel wants to keep apple happy. It's the cost of doing business.

Re: What could I connect this to?

[UnknowingFool]

It was released as a spec just over 2 years ago. It requires a different connector. Considering most motherboards came with LPT ports up until a few years ago and I can't remember using one in almost twenty years, I would say that rapid adoption of new technology may be lacking for some manufacturers.

Re: What could I connect this to?

[viperidaenz]

Legacy support and rapid adoption have nothing in common.

Re:What could I connect this to?

[Anonymous Coward]

That's simply false. There's a large amount of Thunderbolt accessories, including video gear, PCIe expansion chassis (very useful for laptops), and docks. Sonnet just announced this Thunderbolt dock, which seems to be a pretty great deal for laptops.

Looking at the dock they offer (Bluray + 2TB HDD) It's over $500... On what planet does that seem like a pretty good deal?

Re:What could I connect this to?

[putaro]

Unfortunately it doesn't "just work".

I have a Mac Pro 17" with Thunderbolt that I mainly use to hook up an external monitor (Thunderbolt->DVI with a KVM switch in between).

I picked up a LaCie Thunderbolt-SATA adapter to mess with. Plugged it in between the laptop and the KVM switch. Oops. Video quality goes to hell. If I pull the KVM it works better, but that kind of screws up my desktop.

It would have been nice if Apple had put two (or more) Thunderbolt ports on the machine but, hey, all you need is one, right, because it's so fast. Until you get something that doesn't play nice with the spec into your chain.

Thunderbolt is gearing up to be the Firewire of the 21st century (and I say this as someone with a whole rack full of Firewire equipment) - cool but not supported well enough to have any long-term longevity.

Re:What could I connect this to?

[tyrione]

Apple has pissed off all the other CE manufacturers. There will be nothing to plug the other end into. Without general support great features are worthless. Apple is repeating Sony's mistake with betamax. They won't share, thus it will fail. Great technology without support is worthless.

Don't own any professional equipment or work within the NAB world? Do some more research. More and more manufacturers are jumping on-board. https://thunderbolttechnology.net/products

Re:What could I connect this to?

[zugmeister]

RDS field engaging...
zzzzzzzzzzt
Wow! That's so awesome I think I'll take two of them!

Re:What could I connect this to?

[Radagast]

No, Firewire is pretty much dead, although it was good for a while.

It seems to me that Thunderbolt has had faster and wider adoption than Firewire did over the same time after introduction. Thunderbolt is also a lot more useful than Firewire, since it's essentially PCIe over a serial cable. It's fairly trivial to adapt existing PCIe drivers to run the same hardware as an external TB box (or the PCIe card in a TB PCIe chassis), so it's very flexible.

Basically, TB finally delivers on the ancient promise of a universal IO interconnect.

Re: What could I connect this to?

[UnknowingFool]

How is Intel releasing a public spec making it "for Apple"? Thunderbolt solves a lot of different problems. It is the next gen external/PCIe. It is supposed to combine Ethernet, eSATA and USB3. USB 3 which is another Intel product solves other problems. There is overlap between the products. Intel released the spec just over two years ago.

Re: What could I connect this to?

[Anonymous Coward]

Considering most motherboards came with LPT ports up until a few years ago...

PC Desktop users will be the last people to adopt Thunderbolt. One: they have slots and don't really need it. Two: it requires a special graphics card and some hacky loop-back connector.

Once the cost comes down, you could see TB being adopted on ThinkPads and higher-end laptops.

Re:What could I connect this to?

[jo_ham]

Apple has pissed off all the other CE manufacturers. There will be nothing to plug the other end into.

Without general support great features are worthless. Apple is repeating Sony's mistake with betamax. They won't share, thus it will fail.

Great technology without support is worthless.

How are Apple "not sharing"?

The technology is owned by Intel, and was developed with the assistance of Apple. Apple also licenced (for free, in perpetuity) the mini displayport connector and port that TB uses. Intel have been promoting it since launch.

Not really seeing how this is Apple not sharing, given that it's not Apple who actually controls the technology, but such is the way with /. posts - the barest tissue of lies covering the bare bones of the truth. The exclusivity deal was over more than a year ago.

Re: What could I connect this to?

[TheRaven64]

One: they have slots and don't really need it.

That depends on the desktop. Lots of popular cases are small and don't have the space for any more expansion cards. It also depends on the user: how many typical users would prefer to open up the case and install something when they could just plug something into the back?

Two: it requires a special graphics card and some hacky loop-back connector.

Really? The port just exposes some PCIe lanes and the graphics card plugs into the PCIe bus. I thought the standard way of doing it was to stick a PCIe-PCIe bridge chip on the graphics card reserve four of the 16 lanes that the slot provides for the Thunderbolt interface.

Re:What could I connect this to?

[otuz]

Retina MBP's and iMacs have two thunderbolt ports. The rest of the lineup will probably follow soon.

Re: What could I connect this to?

[UnknowingFool]

And the exclusivity wasn't so much an exclusion as it was a huge head start. Since Apple helped Intel, they had working products you could buy the day Intel released specs. Design, engineering, sourcing, and manufacturing would take a competitor at least six months.

Neat

[narcc]

Both users will be really excited about this.

Re: Neat

[UnknowingFool]

You mean the millions of consumers that have Macs recently? Or the ones buying Intel's newest boards?

Re: Neat

[narcc]

The joke, in case you missed it, is that Thunderbolt has ridiculously poor market penetration.

Who cares how many new Intel boards ship with the port, or the tiny fraction of the tiny number of Mac users with a newer unit. There's not much you can actually plug in to the things.

Ask any Mac user if they'd trade the Thunderbolt ports (that they obviously don't use) for some old USB 2 ports.

Thunderbolt, it seems, is the new firewire. Unfortunately for Thunderbolt fans, everyone remembers how firewire turned out and doesn't want get burned again.

Re: Neat

[UnknowingFool]

Except for the fact that it's an Intel technology and that it solves a number of different issues. Tell me how many eSATA devices came out right after it came out? Or did it take years for widespread acceptance. But let's not get facts get in the wayz

Re: Neat

[bored]

Tell me how many eSATA devices came out right after it came out?

Actually a lot, most of them garbage USB/eSATA enclosures for single harddrives sold by chinese manufactures. Probably more than are available now that every x86 motherboard not sold by dell/HP/etc comes with the connectors. Its still the cheapest/fastest way to connect an external drive or two to a PC. A couple of my friends were recently blown away when I compared a little two disk eSATA array I use like a USB harddrive with their USB3 devices. USB3 promises a lot, but seems to sort of suck a lot.

Re: Neat

[Anonymous Coward]

The thunderbolt ports are used alot! Remember, they also work as DisplayPort connectors. I don't think that many Mac users would like to loose the ability to connect an extra monitor for an extra USB port.

Re: Neat

[toruonu]

Being a Mac user I'd happily trade one of the USB ports of my 2010 MBP for a Thunderbolt port for external display daisy chaining. The laptop's still good enough that I've not had the heart to upgrade to a newer one, but Thunderbolt and retina display are the two big items in my next laptop wishlist.

Re: Neat

[vyvepe]

A lot, any SATA drive is eSATA. You just need to add power to it trough a separate cable. They changed only the connector. The mistake was not adding power to the new connector. What a pity!

Re: Neat

[Anonymous Coward]

Power in the connector didn't happen because eSATA was an unofficial hack, something which lived outside the SATA standard for a year or two before the standards committee decided that enough was enough and they had to at least attempt to standardize things so it wasn't just going to be the wild west. There were even competing connectors! And the connector which became the de facto (and later de jure) standard was crap, too, because it was a hack on the shitty SATA connector, whose only merit was cheapness. It certainly wasn't ever designed as a robust external connector; it's barely good enough for internal connections.

The hybrid USB/eSATA jacks are even worse.

That's not all. At the protocol level, SATA was a terrible choice for an external port. People always want to connect multiple devices to one port, and the SATA protocol was not designed for that. It took a crude protocol hack and extra silicon (port multipliers) to retrofit this feature in, complete with poor interoperability because it's a hack. Which is why SATA port multipliers haven't caught on outside a few niches.

SAS (Serial-Attached-SCSI) borrowed SATA electrical signalling, added real (if expensive) connectors to the shitty SATA wafer connectors, and did the protocol right. But it's too expensive for the mass market and will never be adopted in consumer gear.

Basically eSATA must die. It was always a bad idea, a bad hack of a bad base design (as good as SATA is in some ways, it's definitely from the same family tree as ATA, which is to say it's shit). All the success it enjoyed was due to there being no other cheap option for multi-gigabit connections to external disks. Now that there is one (USB3), eSATA will hopefully enjoy the death it richly deserves.

New Standards are nice and all....

[iamwhoiamtoday]

But in the end, it all comes down to cost. Current Thunderbolt displays are rather expensive. Heck, I picked up a dual-link DVI monitor of the same resolution for $275 on ebay! why pay three to four times as much for something with only a small few bells and whistles added on?

Thunderbolt, overall, is great in terms of performance, but it just seems to be well beyond what most folks are willing to pay. It's like that guy who brags about how "My car has a Turbo Kit option from the dealer" but he NEVER SPENDS THE MONEY TO GET IT.

The external drives, the only situation that I'd actually be interested in, are also stupid expensive. In the long run, just better off either using E-SATA, USB3, or internalizing the drives. Same goes for daisy chaining monitors. Want to run tons of monitors? Install more video cards! woo.

no more coffee for me after 5pm, k? ._.

Re:New Standards are nice and all....

[Anonymous Coward]

True cost is something that will matter but I am waiting for the external pcie adapter for TB When I am able to attach a desktop G-Card to my laptop through TB. I will make a switch to only a laptop.

Re: New Standards are nice and all....

[UnknowingFool]

You're looking at this from the narrow viewpoint of a monitor connector. ThunderBolt can replace a laptop dock with a single cable. So instead of hooking up a USB, Ethernet, and monitor cables, just hook up one TB cable. It's not there yet as the specifications were released just over two years ago.

Re: New Standards are nice and all....

[Anonymous Coward]

You're looking at this from the narrow viewpoint of a monitor connector. ThunderBolt can replace a laptop dock with a single cable. So instead of hooking up a USB, Ethernet, and monitor cables, just hook up one TB cable. It's not there yet as the specifications were released just over two years ago.

Have you seen a TB dock for under $200? I'm not being snarky (well just a little) because I want one, and I'm actually curious if you've seen one for a price that isn't insane.

Re:New Standards are nice and all....

[viperidaenz]

You can't connect a 4k monitor to a dual-link DVI connector. You need 4 DVI channels for that.

Re:New Standards are nice and all....

[Sarten-X]

The external drives, the only situation that I'd actually be interested in, are also stupid expensive. In the long run, just better off either using E-SATA, USB3, or internalizing the drives. Same goes for daisy chaining monitors. Want to run tons of monitors? Install more video cards! woo.

I'll get right on that, as soon as I figure out how to stuff more hard drives and more video cards into my smart watch.

The impetus for Thunderbolt isn't to do existing jobs for today's technology. Rather, I get the distinct impression that it's being over-engineered to be the standard of choice for the next few decades. No, we don't need 20 Gbps throughput and 4K video now, but when you walk into a friend's living room and plug your watch into his video game display, it's going to need the bandwidth to stream the real-time-rendered fully-immersive video out and pull in the data from the motion-tracking cameras.

Apple's business model is based on having ecosystems. Apple will make the core systems that everyone will need, and third-party manufacturers will make the accessories. The way I see it, Thunderbolt and the Lightning connector are a part of this. Thunderbolt will be the big backbone for high-throughput and high-quality systems, and Lightning will be the simple connector for miscellaneous (slower) accessories.

Re:New Standards are nice and all....

[jtownatpunk.net]

You can already buy laptops with high end graphics and multiple outputs built in. No need for a clunky external setup. High end gaming laptops can be configured with dual 2gb GTX 680M GPUs or the AMD equivalent. Not exactly anemic.

Re: New Standards are nice and all....

[smash]

Compare a thunderbolt cable to a Cisco 10 gig copper cable and tell me thunderbolt is overpriced.

Re:New Standards are nice and all....

[smash]

Depends what you're doing. If you want to hook up to a fibre-channel SAN or a 10 gig network port (1GBe will get nowhere near saturating your SSD), nothing else will cut it. There already exist use cases for thunderbolt today. They just aren't home-user scenarios.

Re: New Standards are nice and all....

[Mad Merlin]

Compare a thunderbolt cable to a Cisco 10 gig copper cable and tell me thunderbolt is overpriced.

Sure, but even Denon's $500 ethernet cable looks like a great deal compared to Cisco gear.

Re: New Standards are nice and all....

[TheRaven64]

Do you remember the launch of USB? It was released in January 1996, and it wasn't until Windows 95 OSR 2.1 in August 1997 that it was even supported by Windows 95, and then only if you bought a new computer (OSR 2.1 was only available to OEMs, it wasn't an update to existing computers that shipped with USB). It wasn't until around 1998 that there started to be a number of USB mice and keyboards available, and most of those were in translucent plastic to go with the iMacs that Apple had released without legacy I/O - most PC users didn't buy them because they were more expensive than PS/2 or serial ones (lots of Apple users did because the original iMac shipped with the worst-designed mouse in human history). In 2000, I got my first USB-only device (I also owned a joypad that could connect to either USB or a game port, and a keyboard and mouse that were USB or PS/2), and it was a gaming mouse. It wasn't until around 2001 that USB stuff was really ubiquitous, and even in 2001/2002 I remember gamers complaining that you should always use PS/2 mice because of USB latency. It took 5 years for USB to become a ubiquitous connector, and it was starting at the low end, with cheap devices with short(ish) lifespans as the primary use cases. The early devices came with a hefty early-adopter premium.

With Intel pushing it, I expect Thunderbolt will have the same adoption pattern: it will appear in chipsets and become a standard connector on all Intel motherboards before it really becomes widely used, and prices won't drop until economies of scale reach the point where they're selling at least tens and probably hundreds of thousands of Thunderbolt peripherals.

Re: New Standards are nice and all....

[drinkypoo]

You're looking at this from the narrow viewpoint of a monitor connector. ThunderBolt can replace a laptop dock with a single cable. So instead of hooking up a USB, Ethernet, and monitor cables, just hook up one TB cable. It's not there yet as the specifications were released just over two years ago.

So what you're saying is that he's looking at it from the narrow viewpoint of the only currently viable use case? What an insensitive clod!

Re:New Standards are nice and all....

[serviscope_minor]

Current Thunderbolt displays are rather expensive.

Not only that but they're really poor for the money.

As you pointed out:

Heck, I picked up a dual-link DVI monitor of the same resolution for $275 on ebay!

Probably one of theose funny fly-by-night Korean companies. I got one too, albeit a little more expensive. They're very nice.

why pay three to four times as much for something with only a small few bells and whistles added on?

Or bells and whistles taken off...

The thing is that if you buy one of those thunderbolt monitors, they only work with pretty new Macs. You buy the Korean cheap one and it works on anything which can drive dual DVI, meaning any desktop or any machine with a mini or standard DisplayPort socket. Oh and the Korean ones invariably have a VESA mount.

Both only come with one input and one input type, but hey, the Korean one is super cheap.

The pick of the bunch is probably the Dell. It's twice the price of the Korean ones (i.e. half the price of that Mac one) and actually has lots of handy features. It can take in HDMI (not sure if 1.4), VGA (useful since many laptops have dual VGA/HDMI out, and only VGA can drive the full resolution), two DVI ports, DisplayPort and component video.

Re:New Standards are nice and all....

[Princeofcups]

But in the end, it all comes down to cost. Current Thunderbolt displays are rather expensive. Heck, I picked up a dual-link DVI monitor of the same resolution for $275 on ebay!

Because the $275 display looks like shit, kind of like this Dell monitor I have as my second display on my iMac at work. Yes, only high end stuff has Thunderbolt right now. Give it time, and it will be on the eye-numbing low end stuff too.

Re: New Standards are nice and all....

[UnknowingFool]

Only if you haven't been paying attention to the increasing number of professional AV devices coming out. AND you ignore Apple's devices. Like hooking up an HD Apple Retina monitor and having the monitor act as your laptop dock. The thing is, the current equipment coming out is for high-end usage as initial high cost normally is meant for pros. The first DSLRs were thousands of dollars just for the body. Consumers didn't use them; only pros did. These days you can get one for a few hundred that would put the first gen to shame.

There are already several options

[SuperKendall]

What could I connect this to?

Several RAID arrays, gigabit ethernet, multiple monitors, misc external storage (like single disks or a DROBO).

All with one connector...

Yes Thunderbolt stuff was slow to come out, but the rate of arrival has picked up.

Re:There are already several options

[ADRA]

Do each of the devices get their own DMA signalling, or are you crippled to only one device being fast at a time? How does context sharing of the pipe sharing work? I imagine that this -could- be a great step in the right direction, but they need a lot more than just a raw fat pipe to make multiple peripherals fast and responsive.

Re:There are already several options

[nine-times]

Also external graphic cards.

Re:There are already several options

[Mad Merlin]

What could I connect this to?

Several RAID arrays...

I wouldn't suggest it. It'll only take two SSDs to saturate a Thunderbolt bus (or 4 SSDs with 20 Gbit Thunderbolt).

Re:There are already several options

[jo_ham]

Do each of the devices get their own DMA signalling, or are you crippled to only one device being fast at a time? How does context sharing of the pipe sharing work? I imagine that this -could- be a great step in the right direction, but they need a lot more than just a raw fat pipe to make multiple peripherals fast and responsive.

It's basically an external extension of the PCI bus, so works in very much the same way. Instead of slots on the motherboard, it's just rolled into a connector and a cable.

Re:There are already several options

[gl4ss]

Also external graphic cards.

do these now actually exist in feasible, makes sense range that you can actually buy? this was the promise of why the whole thing made any sense at all besides docking your laptop with just one cable..

Re:There are already several options

[nine-times]

do these now actually exist in feasible, makes sense range that you can actually buy?

Total price, maybe not. I've heard of people getting a Sonnet Thunderbolt expansion enclosure, adding a video card that's supported by OSX, and running games. Supposedly it works, but you're talking about hundreds of dollars for the enclosure itself, so it's arguably a better value to just buy a dedicated gaming rig.

It seems to me that it holds a lot of potential. You could theoretically have a very lightweight portable netbook/tablet with a docking station which not only provides additional ports but additional computational power.

But maybe it just appeals to me, and there's not a mass market. I don't like having to keep documents and applications in sync across a bunch of different devices, and I still don't really like trusting "the cloud" to keep it all in sync. So I really like the idea of having a 7" tablet that I can take anywhere, but being able to dock it at home and use it as a full desktop and gaming rig with all the peripherals, without maintaining a separate computer, OS, and profile. We're not there yet, but I think that's where we're going.

Existing stuff is Good Enough

[Gothmolly]

http://en.wikipedia.org/wiki/Principle_of_good_enough

Re:Existing stuff is Good Enough

[smash]

Except for those use cases where it isn't good enough. Which is where thunderbolt is used. People seem to be expecting it to be as ubiquitous as USB or SATA, which is not ever going to happen, because its not a cheap CPU driven consumer-oriented bus.

It exists so that users of portable machines can plug in high speed peripherals. Not single external hard drives, but arrays, fibre channel, external GPUs, etc.

Most users don't do that. And that's fine. But if you DO need to do that, then USB3 just won't cut it.

Re:Existing stuff is Good Enough

[H0p313ss]

People seem to be expecting it to be as ubiquitous as USB or SATA, which is not ever going to happen, because its not a cheap CPU driven consumer-oriented bus.

Never say never. Before the iPad came out everyone and their dog had a field day saying how it would flop. Predicting the future is a tricky thing.

Re:Existing stuff is Good Enough

[Anonymous Coward]

It isn't about saying 'never'. thunderbolt is simply not a technology designed for the end consumer, neither is fibre channel, or rack mounted managed switches or a host of other technologies. that isn't to say the products are a failure, they are simply niche. The ipad predictions were about predicting whether consumers would like or want the product, thunderbolt being ubiquitous is like predicting every person that buys a car is going to purchase a turbo and nitro injector as it will provide maximum performance. it simply won't happen as long as it is an expensive extra, only a small percentage of people can get any benefit out of thunderbolt. People talk about multiple monitors, 4k displays, raid attached storage or chains of high bandwidth devices, this is NOT what consumers do, these are a very small portion of consumers... i.e. a niche.

Re:Existing stuff is Good Enough

[drinkypoo]

It exists so that users of portable machines can plug in high speed peripherals. Not single external hard drives, but arrays, fibre channel, external GPUs, etc.

Most users don't do that.

I feel you have misstated the case by way of apologia. Most portable machines powerful enough to be worth plugging in anything more than a single storage device have more than just one port, so users don't need to plug everything into one port. Using a single cable would be a cool feature, but of the vanishingly few people plugging arrays into laptops, vanishingly few of them need a single cable. Slightly more are willing to pay for the privilege and save a few seconds they weren't going to use anyway, but still not enough to justify the cost of the feature.

The simple truth is that you can build a whole storage server on GigE for less than the cost difference to buy a machine with thunderbolt and external devices. That means it's overpriced, full stop.

Re:Existing stuff is Good Enough

[Anonymous Coward]

Except it's not. The current DVI standards cannot support those resolutions. Hell, Single-Link DVI can't even support 1920x1080 at 120Hz. We won't see high resolution monitors until we can actually run them.

Re:Existing stuff is Good Enough

[smash]

GigE is last decade. Yes the market for 10 GigE on portables is small, but if no one ever builds it the use cases won't ever exist. Thunderbolt likely costs a few dollars to include on a motherboard, the expense is in the transceivers in the cables.

Re:Existing stuff is Good Enough

[smash]

By the time thunderbolt and associated cables are consumer price, USB 4 will be commonplace. And the high end users who need high end throughput will have moved onto thunderbolt 3.

Re:Existing stuff is Good Enough

[Anonymous Coward]

By the time thunderbolt and associated cables are consumer price, USB 4 will be commonplace.

In order for USB 4 to reach the speeds needed, it will need to do all the tricks that make Thunderbolt cables expensive.

Re:Existing stuff is Good Enough

[smash]

It will still be implemented in the manner of the "winModem of buses" with far less hardware assistance.

watch units please

[v1]

Mbps != MBps

Please stop doing that in article summaries. When you start getting up into large numbers like that you can't just expect everyone to "read what you meant to say."

Déjà vu all over again...

[sshir]

Is it me, or it looks like FireWire scenario playing out all over again? Only this time it's not only USB, but also upcoming WiGig to jointly lock it into a small niche...

Re:Déjà vu all over again...

[Anonymous Coward]

It saddens me to see people use "Déjà vu all over again" seriously. It's "déjà vu". That's it. The rest is redundancy that makes it self-deprecating humor.

Re: Déjà vu all over again...

[Anonymous Coward]

It's a Yogi Berra quote.

Re: Déjà vu all over again...

[Anonymous Coward]

Obviously, but dolts like the GGP keep using it as if it were the real usage, not for humor.

Re:Déjà vu all over again...

[AndrewX]

You can experience Déjà vu twice... As in, "Driving through that town an hour ago I felt like I'd seen it before, and now I have that feeling again driving through this other one. It's Déjà vu all over again." It's not redundant because you're actually referring to two separate bouts of Déjà vu.

But yes, the way he used it up there, it should have been just "Déjà vu."

Re:Déjà vu all over again...

[Anonymous Coward]

Is it me, or it looks like FireWire scenario playing out all over again?

Only this time it's not only USB, but also upcoming WiGig to jointly lock it into a small niche...

And this is a bad thing why?

Powerful, expensive stuff for those who need it and are willing to pay for it. Cheap commodity stuff for the rest of us.

Used to be, geeks would be excited by new/fast/powerful/cool things even if they couldn't actually afford them (or had any need for them.) Now they seem to delight in pointing out how 640[whatever] ought to be enough for anybody.

I'm confused...

[JasoninKS]

I'm confused...they're coming out with a second version of a connection that has never really taken off to begin with? Is anyone even using Thunderbolt? I can't imagine the usage rate is very high.

Re:I'm confused...

[Anonymous Coward]

Serious? Thunderbolt has transformed many of the small to medium businesses we deal with. Even the large ones... Say, you wouldn't have need for fiber attached laptops? PCI express docking bays with aja cards? Fast external RAID for small machines (Mac Minis for example) that don't sit in server rooms?

The Usage rate is high in the pro-sumer market, which is what it was aimed for.

Re:I'm confused...

[White Flame]

Well, *something* needs to come out to run higher-bandwidth displays and other peripherals. Even if there's not much using it now, at least once the peripherals become more commonplace there will be connector & driver support available if it goes through all that now.

Re:I'm confused...

[sshir]

And that's the thing - time is running out: Dell already started shipping WiGig docking...

Re:I'm confused...

[Anonymous Coward]

so in other words it is purely a niche product that doesn't have consumer appeal or adoption and is destined to go the way of firewire.

Re:I'm confused...

[Anonymous Coward]

no we just dont buy laptops that suck in the first place and add 100000000 dongles to make up for it

Re:I'm confused...

[Anonymous Coward]

If you don't understand the technical significance of Thunderbolt then you shouldn't be on Slashdot...

Re:I'm confused...

[Anonymous Coward]

of course I understand, Intimately, I spend my life in a datacentre and am familiar with more technologies than you can probably think of. But unlike a lot of other wannabe techs that think because they can use a computer in their mothers basement they understand IT, I also understand that just because a tech has some excellent uses doesn't mean it is destined for blanket consumer adoption. thunderbolt has very little to no benefit whatsoever for the vast majority of consumers, combine that with prohibitive pricing and this is a tech that is likely to be replaced long before it ever reaches widespread adoption.

"20Gb/s" will bring useful 10GbE

[GrumpyOldMan]

I do 10GbE drivers, and the previous generation of tbolt did not really offer 10Gb/s of usable bandwidth to PCIe devices, it was more like 8Gb/s:

If you recall, tbolt muxes PCIe and Display Port. On the PCIe side, the thunderbolt bridge passed 2 lanes of Gen2 PCIe through to devices. Since Gen2 is "5GT/s" per lane, you'd think you'd have 10Gb/s. But not really, as "10Gb/s" does not take into account PCIe overhead, which can be about 20% of the data transfer rate. So on the original "10Gb/s" thunderbolt, you were lucky to get 7Gb/s transfer rate from 10GbE NIC, once you also add in network protocol overheads.

Having a bus-constrained NIC leads to all sorts of weird problems when receiving data.. With flow control disabled in combination with bursty transfers, you often see far less than the 7Gb/s peak, as TCP hunts around to find the constraint and recover from frequent packet loss events.

It sounds like they've built the new part from 2 lanes of Gen3 PCIe, which should be good for ~16Gb/s of usable bandwidth. This is a very welcome change, as 16Gb/s should be enough for a single-port 10GbE NIC running at full speed, and a disk controller talking to a fast SSD or an external RAID array that can deliver ~750MB/s (bytes) of I.O.

Just don't try to use a bonded 2 port 10GbE NIC, or you're back at the bandwidth constrained problem.
 

Re:"20Gb/s" will bring useful 10GbE

[Anonymous Coward]

Just curious. It sounds like you do embedded development. How can I get a job like yours? Do you need domain knowledge in addition to the typical knowledge embeeded devs have?

Apple

[MrEricSir]

Apple has confusingly named their new iDevice connector "Lightning," so I think people can be forgiven for assuming Thunderbolt and Lightning are from the same company.

Re:Apple

[Master Moose]

Thunderbolt and Lightning

Very very Firghtening me!

Re:Apple

[Anonymous Coward]

Thunderbolt and Lightning

Very very Firghtening me!

Galileo!

Re:Apple

[wonkey_monkey]

Galileo.

Re:Apple

[Anonymous Coward]

Galileo.

Re:Apple

[Anonymous Coward]

figaro

Re:Apple

[Anonymous Coward]

Magnifico!

Re: Apple

[Anonymous Coward]

Figaro.

Re:Apple

[Anonymous Coward]

Figaro

Thunderbolt to PCIe expansion for gaming?

[Anonymous Coward]

Is it possible yet?

Business-grade direct-attached storage?

[__aawmso8327]

I sure wouldn't mind seeing this compete with stuff like eSATA and SAS in the DAS market. I know that there is the Promise Pegasus and such - I'm thinking more along the lines of, "shared storage for VMware" than, "fast scratch space for video editing."

Re:Business-grade direct-attached storage?

[Anonymous Coward]

if it is business grade why would you give yourself the limitations of thunderbolt instead many of the cheap higher bandwidth options like mini-SAS that can achieve more than double the performance of thunderbolt?

Where are the data only cards? mac pro* that needs

[Joe_Dragon]

Where are the data only cards? mac pro* that may need some kind of voodoo like loop back cable as the dual xeon systems don't have on board video as part of the cpu / chipset?

Some boards do have on board pci 33 based video mainly server boards.

A video card can max out the bus and be underpowed

[Joe_Dragon]

A video card can max out the bus and be underpowered as Thunderbolt is only like pci-e X4.

a laptop can not replace a workstion system

[Joe_Dragon]

a laptop can not replace a workstation system also the macbook maxes out at 16gb ram and only 1 HDD build in.

Re:a laptop can not replace a workstion system

[otuz]

Which is why you have fast external expansion buses in the laptops to replace the need of having internal expansions. The 10Gbps Thunderbolt is enough for that in many areas, especially if you have several of them. This new version is going to broaden that to applications like high-end gaming and such.

Re:a laptop can not replace a workstion system

[otuz]

Oh and yes, the RAM thing can't be worked around like this yet, but the amount of it grows with each gen until it's "better than enough" for everything but the tiniest niches.

Re:a laptop can not replace a workstion system

[sg_oneill]

I wish apple would realise this. I love mac pros and macbooks ,but they just are not the same sort of product and it feels like apple has forgotten about the humble mac pro.

Re:a laptop can not replace a workstion system

[Anonymous Coward]

I have a 15" rMBP (2.7 GHz i7) at work. It's a capable workstation. Of course, sometimes 8 threads and 16 GB is not enough, but the fix is easy, i.e. ssh onto a super computer.

Re:a laptop can not replace a workstion system

[rjr3]

My desktop and 2 of my laptops are Lenovos. Up until this year my best/favorite devices. OpenSuse 12.3 and Windows 7 - 32bit & 64 bit.

I am writing this on a 3 week old MacPro 2.7 GHz, 512 Gb flash and 16 Gb ram ... and retina display. I'll be hooking it up today to an HP ZR30W so I get 2560x1600 AND 2880x1800 - my 50 year old eyes are in heaven. And I get more IOPs than my VPLEX or Nimble storage can deliver.

It is singularly the best workstation I have ever had in my life. And my experience covers everything from an Intel Z80 running CPM 30 years ago to racks of Sun Enterprise servers on Solaris 10. I used most of the AT&T 3B and the Amdahl stuff too. Lisa, Fat Macs and Multias were cool as well.

wanna dis it - sure, go ahead - you clearly don't own one. But its the best box I've ever had.

Re:a laptop can not replace a workstion system

[amiga3D]

The Mac Pro is a niche system. It's designed for heavy duty multimedia and there is no such thing as enough Ram in that niche. Macbook Pro is good enough for everything else except maybe hardcore gamers that need to upgrade video cards every time a new one comes out.

Re:a laptop can not replace a workstion system

[amiga3D]

nice to have a supercomputer laying around.

Re:a laptop can not replace a workstion system

[Anonymous Coward]

My laptop has 32gb of RAM and dual HDD bays.

and an I7

??

Re:a laptop can not replace a workstion system

[Xeranar]

Or for people that like larger screens, people who don't like apple's OS, or for people who don't think a laptop should cost a month of a mortage. But that's neither here nor there, carry on with your status symbols.

Re:a laptop can not replace a workstion system

[Anonymous Coward]

Or for people that like larger screens

Choose between 3 different sizes, or plug in an external monitor.

people who don't like apple's OS

Install a different OS if you want.

or for people who don't think a laptop should cost a month of a mortage.

They are actually priced right inline with other manufacturers laptops of the same quality. We're talking about business-grade Lenovo, HP, Dell, etc. laptops, not bargain basement Asus.

But that's neither here nor there, carry on with your status symbols.

Ah, here we go. The personal bias comes out in your post, reflecting why you'd post so much misinformation to begin with.

Re:a laptop can not replace a workstion system

[amiga3D]

I know Mac hardware isn't for everyone but really that post just sounds like blatant envy.

Re:a laptop can not replace a workstion system

[otuz]

Or for people that like larger screens

Umm what? I can connect three separate screens directly into my (retina) macbook pro (2x displayport + 1x dvi). More with external thunderbolt PCIe breakout chassis' using more GPU's. One of the external screens I have connected has a resolution of 3840x2400, which is about as big as it gets resolution-wise.

Re:a laptop can not replace a workstion system

[otuz]

Sure, I had a Mac Pro in daily use before I upgraded to the retina mbp. It was absolutely necessary for driving enough displays and a (then) huge amount of 16GB ram and fast enough CPU's. However, I see it as a thing of the past in my case. Technology caught up and provides what I need in a laptop form-factor for the first time ever. Within a few years, the same will happen to most of the remaining workstation users.

Re:a laptop can not replace a workstion system

[Plumpaquatsch]

The Mac Pro is a niche system.

All high-end PCs are niche systems.

Re:a laptop can not replace a workstion system

[Anonymous Coward]

If you need raw performance and OSX, turn a Falcon Northwest DRX into a hackintosh.

Max specs:
dual GeForce 680M 4GB
core i7 3940XM (quad-core 3.00Ghz)
32GB ram
dual 512GB SSDs

Re:a laptop can not replace a workstion system

[Zero__Kelvin]

... but the real question is ... does it have a shift key?

Re:a laptop can not replace a workstion system

[nobodie]

yeah great. I don't fucking want a laptop. I used them from 1994 to 2005, needed them for business, but I don't want to move my computer around, I want to have a desk and a keyboard and a mouse or trackball and everything in one place that doesn't move/can't move. If a thief comes into my house they leave, cause I don't have a laptop and who would try to run out the door with a box? I have nothing )obvious) that is portable and sleep wioth doors unlocked, no guns in the house and windows open wide to the breeze in a transition neighborhood.

Why in hell would i want a laptop when it costs extra money, I have no real control over the hardware, have no desire to move it around and don't want to have someone steal it because I get up from somewhere and go to get a fresh cup of coffee?

Re:a laptop can not replace a workstion system

[nobodie]

What? Mac hardware is not "business grade" and never has been. Mac marketing IS business grade and always has been, yes that is true and they deserve the cred for what they have achieved in that realm. To support that marketing, they provide shiny and pretty, obviously different designs based on jon Ives vision of what is interesting, appealing and different. It has worked. But the quality? Not so much. What they do is this: They contract with a manufacturer to provide a certain number of units of a part at a specific price (just like Acer or Asus might do for something they don't manufacture themselves), they spec that part in the mid-range of consumer gear (just like Asus and Acer might do for a mid-range product) they sign an agreement that this part, with these specs (especially the design aspects that fit into the overall design, like the exact dimensions, where the wires port, how they ground, the access port and other things I'm failing to think of and then they have it made, shipped to their final assembly location and then tested (just like everyone does) and shipped to you or the Apple store.
Except for the marketing , the parts themselves are nothing special, certainly not reliably business grade. They do some things (for marketing purposes) like thunderbolt that also help them to differentiate their products. One of the best things they do is to dead-tech stuff faster than most manufacturers, which is a good thing IMHO. But "business grade", not so much.

Re:a laptop can not replace a workstion system

[otuz]

Maybe you quit them a bit too early? I've also had laptops roughly the same time you've had, but they were basically peripherials to my desktop machines, never a replacements like they became recently.

next generation of graphics cards will want pci-e

[Joe_Dragon]

next generation of graphics cards will want pci-e at least a X16 2.0 link or a X8 3.0 one. thunderbolt is way under that it's not even pci-e 2.0 X8

Why not just use 10GbE?

[enos]

Do you have any insight why they even bother with TB when 10Gb Ethernet already exists and has been deployed for ages? I.e. why not just use 10GbE instead?
It seems like reinventing the wheel for no real gain.

Re:Why not just use 10GbE?

[Strider-]

Do you have any insight why they even bother with TB when 10Gb Ethernet already exists and has been deployed for ages? I.e. why not just use 10GbE instead?
It seems like reinventing the wheel for no real gain.

When all you have is a hammer...

The main reason for using Thunderbolt over 10Gb Ethernet is that one has a fairly significant protocol overhead (Ethernet) while the other is primarily a bus protocol, and operates at a much lower level than Ethernet does. Each has their strengths and weaknesses, each has their application.

Re:Why not just use 10GbE?

[smash]

Because 10GBe doesn't expose PCI to your peripherals. Thunderbolt isn't JUST for 10GBe NICs, although that is a popular high bandwidth application that no other external connector can currently provide.

Re:Why not just use 10GbE?

[TheSync]

Because 10GBe doesn't expose PCI to your peripherals

How about ExpEther technology virtualizes PCI Express over Ethernet.

Re:Why not just use 10GbE?

[Blaskowicz]

You gotta plug that 10GbE controller somewhere, right?
Currently the bus of choice is PCIe, with differing physical forms : chip soldered onboard, regular card, expresscard (similar to PCMCIA) and then TB.
If your PC doesn't have 10GbE onboard and no available PCIe 4x, 8x or 16x slot, but has a TB port then you gotta use it if you want to "just use 10GbE".

Re:Why not just use 10GbE?

[Blaskowicz]

sorry I must have missed you.
Well, 10GbE is even more expensive than TB and uses more power, which makes it not that desirable on a consumer laptop. Plus you can use TB for something other than storage and networking.

Re:Why not just use 10GbE?

[Blaskowicz]

misread

Re:Why not just use 10GbE?

[smash]

Whilst that's neat... it's no doubt much less less efficient than just exposing PCIe in the first place without going from PCIe->Encoding -> Ethernet -> Decoding -> PCIe.

using pci-e 3.0 on the Qm77 chipset will cut video

[Joe_Dragon]

using pci-e 3.0 on the Qm77 chipset for Thunderbolt will cut video down to X8 3.0 and some boards may use switchers to get full use out of the pci-e lanes.

http://www.intel.com/content/www/us/en/chipsets/performance-chipsets/mobile-chipset-qm77.html

but only 2 lanes on the TH side still makes it useless for video cards.

Cables / connectors.

[fahrbot-bot]

... Thunderbolt tech enables 4K video file transfer and display simultaneously in addition to running at 20 Gbps. It will be backward-compatible with previous-gen Thunderbolt cables and connectors ...

And even faster with gold-plated Monster cables / connectors !

Re:Cables / connectors.

[Anonymous Coward]

But only if they have arrows to tell you which way the signal should flow (aka which way the money should flow when separated from a fool)

Re:Cables / connectors.

[EmagGeek]

Monster cables are crap. I buy only high-quality digital cables, and can't forget the power conditioner to remove "noise pollution" from my 60Hz AC power to "reveal hidden subtleties of the entire audio event as it eliminates the widespread negative effects of high frequency noise pollution" *bangs head on desk*

Macbook vs Mac Pro

[fyngyrz]

But imho the mac pro is obsolete and has been replaced by a macbook with a thunderbolt port.

If you wanted to configure your macbook to match a *current* mac pro, you'd need 8 more full i7 cores (assuming you have four in the macbook), four hard drives, four external graphics engines, and 48 gb (I think) of RAM... all strung out on your thunderbolt cable. And a *lot* of power supply wiring. Not sure that's an equivalence that is worth much.

And add to that whatever they do with the next Mac pro upgrade they say they're working on... More cores? More ram? Faster system bus? All of the above? No, don't think your macbook is quite there, lol.

Re:Macbook vs Mac Pro

[TheRaven64]

It depends a lot on what you're doing. What do people do on Macs Pro that are that processor intensive these days? I do a fair bit of compute-bound work (big compile jobs, FPGA place-and-route, and so on), but I do it sitting in front of my MacBook Pro and ssh'd into some big servers each with 32 cores and 256GB of RAM. The current Mac Pro maxes out at 12 cores and 64GB of RAM, and for that price costs more than we paid for the servers (which can be easily shared between people, as we all tend to have bursty workloads). Interactive video / image processing tasks tend to be offloaded to the GPU these days and there isn't such a huge difference between the desktop and mobile parts there. Bigger jobs are best run on a load of cheap commodity hardware in a rack than on a workstation.

Re:Macbook vs Mac Pro

[fyngyrz]

What do people do on Macs Pro that are that processor intensive these days?

Speaking for myself, 48-bit image processing in a layer-based, non-destructive paradigm. Software defined radio -- extremely demanding, that. High speed data, maximized low latency requirements, no particular limit to the amount of processing one might like to apply to the signals / spectrum segment. I use Logic for musical performance, and that's absolutely got to stay local, again latency must be managed to then nth degree and the more processing that can be done within that bound, the better. None of this can be handed off down the network; it just wouldn't work well. Or at all.

I sure would like to see core-per-file parallel compilation, too, but instead, all of the dev environments I use keep the source and object on HD and do them one at a time, serially. Big projects take much longer than the hardware at hand could manage. XCode, QT, gcc/gmake... all serial sluggards.

The current Mac Pro maxes out at 12 cores and 64GB of RAM

Well, sort of. My understanding (which could be wrong) is that those 12 cores are all hyperthreaded, so as long as you're not requiring FPU or blowing cache a lot, you're more-or-less running 24 cores. Not sure as I have an older generation dual four-core w/o hyperthreads (but with two sets of four FPUs.) It's a fair bit of computing power; my Macbook pro, a dual core with lots of resources, can't even come close to keeping up with my older Mac pro.

Interactive video / image processing tasks tend to be offloaded to the GPU these days

That varies. We use an integer-math approach that maximizes CPU power and doesn't rely so much on the highly variable GPU capabilities of the various Mac models (not to mention having to drink more of the Apple kool-aid than really needed... we really like portable code.) There are a number of advantages to this, first among them the availability of a great deal more RAM (it's useful to stack a hundred or so 48-bit astro images, for instance) and so more layers-per-image, but also a more consistent performance for the end user. Some Macs -- the current minis, for example -- use bottom feeder Intel shared RAM GPUs. They aren't anything to write home about. Even so, the machine can be carrying a 2.6GHz i7 w/16gb. We often outperform Aperture, Apple's poster child for GPU use in image processing; can't really say why, but there it is.

Bigger jobs are best run on a load of cheap commodity hardware in a rack than on a workstation.

For some value of bigger jobs, sure. For the things that I do on my desk, no.

In any case, a Macbook isn't going to provide even close to the same level of hardware as a Mac pro, which was really my only point above.

Re:Macbook vs Mac Pro

[klubar]

Other than the looks, you don't really need the MBP sitting on your desk. You could use any commodity laptop to SSH into the server farm. Nothing wrong with the MBP, but overkill as a dumb terminal.

Re:Macbook vs Mac Pro

[fyngyrz]

Ok, so we've established that you don't know anything about SDR software (tip: the desktop software is not doing the the same tasks as the hardware, and there is a huge amount more that can be done once the baseband data is in... dual core laptops don't have the power required to go very far down that road... and yes, I'm the author of one of the more sophisticated SDR packages out there); you don't understand what is required to make multiple compiles work *well* (tip, it's not just an option to make), you don't understand opencl's limits at *all*, nor what happens when threads collide over FPU demand, you misinterpret one function (stacking) as "specialist tool" (exact same kind of tool... that's why I compared them, the stacking is one function out of many and was brought up only to point out a use for many layers, which you promptly, again, showed you didn't understand), you don't understand what the fact that *basic* use of Logic leaves CPU to spare, which means that *advanced* use of Logic has CPU remaining to use, *and can use it*, and can take latency out of the acceptable range quite easily -- the more power you have, the more elbow room you have before that happens. In short, you're moderately to wildly confused about just about everything you wrote about.

Well, that was fun. Cheers. :)

Re:Macbook vs Mac Pro

[Maxoverdrive]

I sure would like to see core-per-file parallel compilation, too, but instead, all of the dev environments I use keep the source and object on HD and do them one at a time, serially. Big projects take much longer than the hardware at hand could manage. XCode, QT, gcc/gmake... all serial sluggards.

have you tried -j?

Re:Macbook vs Mac Pro

[Anonymous Coward]

Ok, so we've established that you don't know anything about SDR software

Conceded!

you don't understand what is required to make multiple compiles work *well* (tip, it's not just an option to make),

Spare me. I see I have to remind you what you wrote:

I sure would like to see core-per-file parallel compilation, too, but instead, all of the dev environments I use keep the source and object on HD and do them one at a time, serially. Big projects take much longer than the hardware at hand could manage. XCode, QT, gcc/gmake... all serial sluggards.

When called on this, your comeback is 'butbutbut you don't know how to do it right ', a transparent ad hominem fallacy. Worse, one based on nothing but your imagination, as I said nothing about the details. Kindly fuck off.

Before you try to deflect with more nitpicking, yes, I am perfectly aware that gcc itself is limited to one file at a time. The issue is that you said there's no parallelism available at all, even though you listed multiple things which implement it by spawning many copies of gcc (or other compilers). XCode and GNU make do this for sure, probably QT as well.

For crying out loud, man, you don't even have to know anything to get XCode to do parallel builds! The GUI does it for you when you press the big button.

you don't understand opencl's limits at *all*,

So, suggesting that you do image processing in OpenCL, something it's pretty damn well suited for, implies I know nothing about its limits? Sure, if that makes you feel better, o wise one.

nor what happens when threads collide over FPU demand,

I would be fascinated to know how you decided I don't know anything about this topic, as it's not something I commented on.

Here, I'll fill the void so you actually have something to hang your criticism on. It sounds like you've got at least a dim understanding of why programs with a high percentage of FP instructions often don't benefit much (or at all) from Intel's hyperthreading feature. Non-controversial to anyone who understands how Intel HT works. It simultaneously issues instructions from two threads to one CPU core, taking advantage of the fact that one thread usually can't fully utilize all execution units (EUs) on its own. However, many (not all) FP programs are capable of using all the execution slots of a single core's FP EUs without assistance from a second thread, so there's no HT performance gain when adding a second FP heavy thread.

(You could get a nice benefit from pairing a FP thread with an integer thread, but so far as I know nobody has seriously tried to make this happen reliably in general purpose systems since it's difficult for the OS process scheduler to know which threads have what instruction mix. And that would only work if there was an integer thread to run, of course.)

None of that has anything to do with the fact that you're really dumb for apparently insisting on scalar integer math for image processing, which is a horrible way to go if you want to maximize performance.

you misinterpret one function (stacking) as "specialist tool" (exact same kind of tool... that's why I compared them, the stacking is one function out of many and was brought up only to point out a use for many layers, which you promptly, again, showed you didn't understand),

It's astonishing how much wrongheaded misreading you do. I said your tool is a specialist tool because you mentioned 48-bit color depth and astrophotography. Of course it's the same general kind of app as Aperture. But in every application, some features are important just to have in the toolbox while others need to be carefully honed, and the choices about which tools fall into what category differ based on the target audience. Call it a hunch, but I'm guessing that your

Re:A video card can max out the bus and be underpo

[smash]

Sure, it won't be as fast as the same card running in a desktop. But it will be FAR quicker than any mobile GPU available, even running on PCIe x4. Why? Because even at PCIe x16, transfers across the PCIe bus are slow. This is why cards have onboard memory.

USB 3.0 has already won

[Anonymous Coward]

USB 3.0 is backward compatible with prior versions and is fast enough for most of the stuff people use. They're has to be a killer application for Thunderbolt for mass adoption, but that has yet to appear. Combine that with ridiculous prices for Thunderbolt cables and it looks like a failure for wide adoption.

Re: USB 3.0 has already won

[UnknowingFool]

There is no "winning" between USB 3 and TB as they are designed for different purposes. Sure there is some overlap in their usage. It was the same as the FireWire vs USB debate. USB was cheaper and more prevalent. Yes you could use it to transfer gigabytes of data to/from your computer. However if you needed it to do it all the time (like in HD video editing), it sucked. Prosumers used FireWire which cost more for this reason.

It's the same reason Black and Decker is probably the number #1 power tool company in the US. It's also the same reason B&D tools are rarely used by pros. If you need some power tools around the house now and then it's adequate. If you use it everyday, you want something better suited to the job.

All I see is 2 x high powered USB3.0 ports...

[sidevans]

Thunderbolt...
-------------------
Voltage : 18v
Current : 550ma
Pins : Total of 20
6 x Ground Pins
4 x Data (+)
4 x Data (-)
6 x Other stuff
Power = 10 watt

USB3.0 (Powered)
-----------------------
Voltage : 5v
Current : 900ma
Pins : Total of 11
3 x Ground
2 x Data (+)
2 x Data (-)
4 x Other Stuff
Power = 5 watt

SO basically, they've increase the voltage by 400%, dropped the current by half, added some more wires and made some special new connectors? Oh and given it another name which is $Nature_Scary_Thing . $Electricity_Word to match their previous hipster names? I'm not seeing the big deal here yet...

Re:All I see is 2 x high powered USB3.0 ports...

[aristotle-dude]

SO basically, they've increase the voltage by 400%, dropped the current by half, added some more wires and made some special new connectors? Oh and given it another name which is Nature_Scary_Thing . Electricity_Word to match their previous hipster names? I'm not seeing the big deal here yet.

Then you need to get your eyes checked. It also offers display port pass through and access to the PCI bus. Does USB3 offer any of that? Oh, and you can get USB3 and Gigabit Ethernet through a Thunderbolt dock if you really want to.

In a nutshell, USB3 can be a subset of Thunderbolt but not the other way around.

Re:All I see is 2 x high powered USB3.0 ports...

[gl4ss]

Oh I see, instead of buying a laptop with HDMI / Gb Network and USB3, I can buy a laptop with a single ThunderBolt port and also buy a whole bunch of adapters or a dock to carry around in my laptop case?

it's the apple way! modders be hatin'.

Re:All I see is 2 x high powered USB3.0 ports...

[aristotle-dude]

Oh I see, instead of buying a laptop with HDMI / Gb Network and USB3, I can buy a laptop with a single ThunderBolt port and also buy a whole bunch of adapters or a dock to carry around in my laptop case?

My rMBP laptop has an HDMI port, SD card slot, two USB 3 ports and two thunderbolt ports. Each of those thunderbolt ports can break out Gigabit Ethernet, additional USB 3 ports, display port, Firewire 800 and passthrough thunderbolt with the right dock/breakout box.

With Thunderbolt, you can daisy chain a thunderbolt display and thunderbolt storage off one of those ports.

Like I said, it already has HDMI out but if I need Firewire 800, I can use an adapter. If I need Ethernet, I can use an adapter too but I don't use ethernet at home and most places have wifi these days. I don't always need Firewire 800 or ethernet but if I do, I can plug in the adapter for those rare times that I do.

Re:All I see is 2 x high powered USB3.0 ports...

[Anonymous Coward]

You're really kinda dumb, you know? I'll give you a hint: Thunderbolt supports a total of about 19 Gbps. USB3 supports about 4. And even without the speed difference, Thunderbolt does things that USB can't do, because the USB protocol doesn't permit it, having been designed 15ish years ago for things that are really slow by modern standards. It's not just about fucking with voltages and connector pincounts, dumbass.

Also, you need to learn how to count. USB3 has 9 pins, not 11, and it's 2 grounds, 3 data (+), 3 data (-), and 1 power. (3 data pairs because you have one bidirectional pair for legacy USB2.x/1.x, and for USB3 mode, one transmit pair and one receive pair.) I happen to know that offhand, didn't bother to check Thunderbolt, but I'm sure you got it wrong somehow too.

Re:next generation of graphics cards will want pci

[Chrontius]

I'm throwing science at the wall to see what sticks.

I actually wouldn't be entirely surprised if the oil-immersion GPU happens eventually, though.

Still not good enough!

[bertok]

Why are manufacturers coming out almost-but-not-good-enough connector standards one after another?

Both tablets and TVs are leaving PC displays in the dust, and new PC connector standards that aren't even available yet already don't have the required bandwidth to support displays that are coming to market now, let alone in the future!

For example, support for full 4K video over 20 Gbps is bullshit, because some aspect of the full spec has to be abandoned:

Resolution: 3840 x 2160
Bits per pixel: 30 or 36 (10 or 12 bits per color channel)
3D or High Framerate: 120 fps

This adds up to: 3840 * 2160 * 30 * 120 = 29.8 Gbps.

Sure, you can drop the framerates, but then expect to have a headache viewing 3D. The bit-depth can be lowered, but then expect visible banding when using gamuts that are wider than sRGB. The resolution can't be lowered, because calling 3840 pixels "4K" is already a stretch.

Re:Still not good enough!

[advid.net]

3D or High Framerate: 120 fps

Sure, you can drop the framerates, but then expect to have a headache viewing 3D.

60 fps only are required for "3D" at 30fps

72 fps is the maximum rate that enhances user experience, above the eyes and brain don't feel there's a difference (for 2D display)

20 Gbps are enough for this.

Re:Still not good enough!

[Anonymous Coward]

Wait, "4K" stands for vertical pixels? Seriously? ... My dreams are crushed. I was way more excited about this than I should have been. :(

Re:Still not good enough!

[maccodemonkey]

3D or High Framerate: 120 fps

Huh? Most video is at 24 fps. Even if we generously triple that, we're nowhere near 120 fps. Hobbit 3D's highest encode was 48 fps which was considered super high quality, and most theaters were still at 24.

Re:Still not good enough!

[AndrewX]

72 fps is the maximum rate that enhances user experience, above the eyes and brain don't feel there's a difference

I love how some people tout little 'facts' like these (some people say it's 30, some people say it's 60), yet I and almost everyone I know can clearly tell the difference between 72 and 120. Don't try to bring the rest of the world down just because your eyes are slow.

Re:Still not good enough!

[sexconker]

3D or High Framerate: 120 fps

Huh? Most video is at 24 fps. Even if we generously triple that, we're nowhere near 120 fps. Hobbit 3D's highest encode was 48 fps which was considered super high quality, and most theaters were still at 24.

24 fps is trash.

30 fps is trash.

48 fps is trash too. If you want the Hobbit in decent 3D you need to display 96 fps. If you want to watch it on a fixed-refresh rate display (all modern consumer TVs and monitors) you're fucked. We just got 120 Hz displays which allow near-perfect display of 24 and 30 fps content. With the Hobbit in 2D you'd need a 240 Hz display to properly handle 24, 30, and 48 fps content. Want decent 3D? Now you need 480 Hz. The dipshit should have gone with 60 Hz. Then 120 Hz displays would display it just fine, even in 3D. The BluRay would display it just fine as well, though at a culled frame rate. Broadcasters would have plenty of options for stuffing the thing into their standard 60i streams or 60p streams as 2D or 3D.

I really just wish the fucking world would settle on 60 fps.

Re:Still not good enough!

[Tapewolf]

I really just wish the fucking world would settle on 60 fps.

There is a slight problem with that:

http://en.wikipedia.org/wiki/File:PAL-NTSC-SECAM.svg

...the green bits are 30fps. Everything else, i.e about 2/3 of the planet, is 25fps. 25 does not go into 60. I suppose at a pinch it could be run at 24fps with the 120Hz timebase you suggested, but that's not exactly optimal...

Re:Still not good enough!

[sexconker]

I really just wish the fucking world would settle on 60 fps.

There is a slight problem with that:

http://en.wikipedia.org/wiki/File:PAL-NTSC-SECAM.svg

...the green bits are 30fps. Everything else, i.e about 2/3 of the planet, is 25fps. 25 does not go into 60. I suppose at a pinch it could be run at 24fps with the 120Hz timebase you suggested, but that's not exactly optimal...

The trouble is that your "2/3 of the planet" equates to about "10% of produced content". Besides, the people in the land of PAL have had to deal with 24 to 25 fps horse shit since the dawn of time, so they're used to it. And dropping 1 out of every 6 frames to go from 60 to 50 is much less offensive than the 24 to 30 shit we deal with now. And PAL itself is going down the toilet in favor of DVB, so there are no analog constraints even for broadcast signals that restrict anyone to 50 fps.

60 Hz (120 for 3D) for production would solve everything. Broadcast/cable/satellite can push it out whenever the fuck they please as long as they're using digital. 24 fps, 30 fps, 60 fps fit perfectly. 25 and 50 fps shit will have less distortion than anything that's ever been telecined. People can display it to the best of their TVs ability. And for NTSC shit (the vast majority of content) that PAL people want to watch, you could send the original frame rate shit to the PAL territories and have it display properly. No need to run a film at sped up at 25 fps or send video 5-in-6 decimated (30 to 25).

Re:Still not good enough!

[Blaskowicz]

3D is made of two frames, one for each eye, so to display 48 fps 3D stereo you usually need as much bandwith as for displaying 96 fps.

Re:Still not good enough!

[advid.net]

This isn't about "some people say" or my eyes, this is the result of an experiment, which conclude that 72 fps is the maximum rate that enhances user experience.

See http://en.wikipedia.org/wiki/Frame_rate, about 72p (yes, citation needed, I know).

If you maid an experiment (double blind conditions) that shows that you can tell the difference between 72fdp and 120fps, you should report it, please.

The display integration thing has hurt it

[Sycraft-fu]

Originally Light Peak was supposed to basically just be an external PCIe bus (and it could be internal too). The idea was a connector for things that need lower overhead than USB, and also hopefully eventually a single connector for all kinds of things. With the original goal of 100gbps, that would have been realistic (optical was the original interface design).

However things got changed pretty quick in part for cost reasons, but also because Apple got involved (meaning gave Intel money). Apple is obsessed with less cables because cables = evil in their mindset. So it got changed to be display + PCIe on one cable.

That had negative implications for the bandwidth, but also for the cost and ability to implement it. If it was just PCIe, well then a PCIe-thunderbolt card would be real feasible, and you could add a thunderbolt port by hanging it off the PCI bus. However with display integrated, it needs to work with the integrated display adapter and all that jazz.

Ultimately more cost, and thus less interest. While some Apple types might salivate over the prospect of one cable that goes from a laptop to a monitor, and then a bunch of non-monitors ports on that monitor, most people don't care.

Re:The display integration thing has hurt it

[Electricity Likes Me]

I don't get the one cable thing. Cable density has never really been the problem - it's connector variety. If I need 10 cables, but they're all the exact same connector and can be plugged in wherever, then who cares how many there are - it takes no effort to manage them.

A laptop with 10 Thunderbolt ports along the back would have tons of expandability, you'd never be port starved (maybe bandwidth starved) and they could all be easily hidden with jack covers or whatever.

Re: The display integration thing has hurt it

[UnknowingFool]

If you connect/disconnect your laptop many times a day, how many cables would you like to handle. Most laptops these days have docks but they are not universal between manufacturers much less with the same manufacturer. Now with a company laptop, my company foots the bill. A consumer doesn't want to buy a new dock when they change laptops. If they all used TB, then it is much more convenient.

Thunderbolt devices

[theVarangian]

note: jedidiah is a gnu/hippie who's angry that Apple took-over the *nix desktop market.

In any case, Thunderbolt has been out for two years now and the peripheral selection is pretty pathetic. Apple has an expensive monitor/docking station. Belkin's docking station has been "coming soon" forever now. And there's some drive enclosures, and that's abou tit.

The Belkin and Matrox docks have been out for a while now. The Belkin dock was "Temporarily out of stock" at Amazon despite the $299 price tag the last time I checked I checked (about 60 seconds ago). There are no drive enclosures available (unfortunately), only ready made external drives which only make sense if you have an SSD to really take advantage of Thunderbolt's speed and a capacity of at least 128gb which makes them extra expensive. Buying a Thunderbolt enabled mechanical drive only makes sense if you are stuck with a machine that has a Thunderbolt connector and USB2 connectors since you get no speed advantage to speak of over USB3. For my purposes external drives start getting interesting at 500gb since I use them mostly for backups and to store tons of photographs, Photoshop files and e-books. There is also a SATA adapter from Seagate (Although strangely enough no SSD disks), a string of really interesting and ridiculously expensive RAID solutions and a few adapters for FW800, GigaBit ethernet., Video etc... What is keeping Thunderbolt down is the price of high capacity SSDs and the fact you can't get any empty enclosures. As the price of SSDs starts to fall and mechanical disks go the way of the dodo (not shedding any tears) and Thunderbolt stands a good chance to compete with USB3 as long as it can keep a speed advantage because I'll buy what ever gets me as close to native SSD+SATA speeds as possible when I'm making large data transfers. Another thing to consider with Thunderbolt is that the display connector usually doubles as a Thunderbolt connector. Since some machines like the MacBook AIr only have one connector you'll either have to unplug the display every time you want to plug in a Thunderbolt device or put your display on the end of the daisy chain so never buy a TB device that does not support daisy chaining. The other option is to buy a machine with two TB ports like the MacBook Pro which is what I did. It's only marginally heavier than a MBA and has way more connector options.

Re:Thunderbolt devices

[Anonymous Coward]

mechanical disks go the way of the dodo

Lol, pfft. I was shopping recently in Akihabara. Just to pick a random example, in one shop the biggest available SSD was 256MB. The biggest real disk was 3TB.

256 *megabytes*? Are you kidding me?

SSD isn't going to take over from real disks unless and until they can make it bigger AND cheaper, and since real disks keep getting bigger, that doesn't look likely to happen any time soon.

(not shedding any tears)

Whilst you're welcome to your own opinions, I personally won't be storing anything of any importance on an SSD until not only are they a viable option cost-wise cf. real disks, but also there's some hope of recovering data off them when they fail. Real disks are well-understood in terms of data-recovery, whereas all SSDs store data in extremely-proprietary ways.

Re:Thunderbolt devices

[Electricity Likes Me]

What's keeping Thunderbolt down is the lack of widespread roll out and affordable PCI-E x16 enclosures to use with graphics cards.

Paying $800 so I can use a regular graphics card with my laptop is absurd. For that price I can buy the entire computer and GPU I'd need in a desktop format.

And the only reason it's expensive is because there's just no volume.

Re:Thunderbolt devices

[amiga3D]

I agree about storage. I bought a 3TB external drive on black friday for 99 dollars. I also bought a 240GB SSD drive for 155 dollars. I put the SSD on the laptop for the OS and it's Applications to use and hooked the 3TB external to it for all the stuff I just want to keep. Best of both worlds.

Re:Thunderbolt devices

[amiga3D]

Thunderbolt will never be cheap. Even now Firewire devices cost more than their USB equivalents and so will it be with Thunderbolt. People use USB because it's good enough and those that use Thunderbolt are those that need the speed and are willing to pay the price.

Re:Thunderbolt devices

[Electricity Likes Me]

Cheap maybe not, but 4-5x the cost of the graphics card you might want to use is absurd for an interface.

Re:Thunderbolt devices

[theVarangian]

mechanical disks go the way of the dodo

Lol, pfft. I was shopping recently in Akihabara. Just to pick a random example, in one shop the biggest available SSD was 256MB. The biggest real disk was 3TB.

256 *megabytes*? Are you kidding me?

SSD isn't going to take over from real disks unless and until they can make it bigger AND cheaper, and since real disks keep getting bigger, that doesn't look likely to happen any time soon.

Really? I've had SSDs in my machines for the last 3 years and I've never cast a nostalgic backward glance. My current laptop drive is a 480Gb SSD and I'm not going back to mechanical any time soon. If you really think mechanical disks are going to trump SSDs you should get together with the people that told me 15-20 years ago that digital photography would never push film cameras off the stage. I dunno about what pros have done in terms of digital vs film cameras but It's been a loooooooooooooooooooong time since I last saw a person with a film camera.

(not shedding any tears)

Whilst you're welcome to your own opinions, I personally won't be storing anything of any importance on an SSD until not only are they a viable option cost-wise cf. real disks, but also there's some hope of recovering data off them when they fail. Real disks are well-understood in terms of data-recovery, whereas all SSDs store data in extremely-proprietary ways.

Ditto :-) If data recovery after a catastrophic hard drive failure is your preferred backup strategy they you have a point. Myself I do regular backups to external storage so It's not an issue for me.

Re:Thunderbolt devices

[Anonymous Coward]

According to Belkin.com, the thunderbolt dock is not yet available.

--AC you replied to.

Re:Thunderbolt devices

[Man Eating Duck]

Whilst you're welcome to your own opinions, I personally won't be storing anything of any importance on an SSD until not only are they a viable option cost-wise cf. real disks, but also there's some hope of recovering data off them when they fail. Real disks are well-understood in terms of data-recovery, whereas all SSDs store data in extremely-proprietary ways.

A curious backup strategy indeed. I agree that an SSD is overkill for bulk storage of movies and similar data, but if you tried putting your OS partition on one you'd never go back. A drive crash will never catch you unaware unless you are computer illiterate. I would never trust or even bother with drive recovery, as it should never be necessary if you have any idea what you're doing.

You might be joking, but: there many relatively cheap or free solutions for both near-real-time backup and availability, which allows you to enjoy the speed of an SSD without worry. You could even go for two drives in a RAID 1 config if availability is very important to you. If you have a laptop with no extra internal bays insert a thumb drive and script periodic rsyncs (with --link-dest) of important documents to that. I'm happy with using Crashplan with ample redundancy (including cloud) for all important data, and doing a full dd of my OS drive regularly to save time in case of recovery. dd is a lot faster than re-installs, and since my OS drive is tiny in comparison to my bulk storage drives, space is not an issue.

My SSD with OS partitions on it has 17842 hours of operation on it now, it has perfect health according to SMART data. It will likely last longer than any other component in my main computer. On the other hand my tertiary drive (seagate 1TB) were retired two days ago after 24040 hours total, with a couple of weeks of illness at the end (wohoo, free rare earth magnets). I had plenty of time to synchronise the remaining content onto the secondary in advance, SMART is your friend. If the SSD suddenly dies tomorrow I'll be back in business in an hour after getting a new one from the LCS. The one component I will never forgo when building a box with a GUI is an SSD for the OS partitions.

A neat trick for gaming, btw, is to install your games to a regular HDD, and move the games you're playing at the moment to SSD (mklink or ln -s manually if you want full control). "Swapping in" a new game literally (yes, literally) takes something like three minutes, and a 120GB SSD will then suffice for your OS and a few games at a time. It truly works wonders for performance.

Woa. This post kept building itself up, and to boot it was a reply to AC who will probably never see it. I'll just leave it here :)

Only if you are ok with them being ass-slow

[Sycraft-fu]

Thunderbolt has 2 lanes of PCIe 2.0 (this new version changes that to 3.0). 10gbps raw data rate, around 8gbps effective. It also has one channel of DisplayPort 1.1a.

So in terms of non-display devices, that means one RAID array of reasonably fast drives can easily overload it. I dunno about you, my RAID controllers usually hand of of 4-8x slots. 1 good SSD can kill half of that on its own. A 10 gig NIC is more than it can handle (look in the thread for a post by someone who implements those). In terms of display, DP 1.1a has enough bandwidth to get you 2560x1600@60fps. Knock on a second display at that rez? Well you don't have enough bandwidth anymore, so you are going to have to reduce rez, or framerate.

Or you could always, you know, have more than one connector and not bitch.

Seriously the one connector thing seems a little silly to me. A marketing solution looking for a problem. Yes, it'll work fine for the kind of stuff Apple likes to do: A laptop connected to a monitor, which then provides USB ports n' such, all over one connector. Ya. Great. Not really that big a deal.

It is not something, at least at present, that you can effectively hang a bunch of shit on one connector and get high performance.

Same argument for USB3

[DrYak]

People buying 4TB or soon larger external drives are not going to be so happy waiting for them to fill up with movies or whatever at USB speeds

Same argument works for USB3. People might as well turn to USB3 (specially since there's NOT A SINGLE OTHER VENDOR[*] beside Apple selling laptops with Thunderbold), so may fork out for a USB3 ExpressCard and go for USB3.

And given how back then people simply prefered moving from USB1 to USB2, instead of moving to more expensive and rarer (although better) FireWire, chance are high that people will move to the cheaper, more available and bettre back-compatible USB3 than move to an exensive and (currently) Apple-only solution (even if technically, a connector with combined displayport and PCI-e is a whole level of awesomeness better and more interesting).

But overall USB1->USB2->USB3 seems to be the path of least resistence (back-compatibility, cheaper, more available).

Note: Even if I'm not even interested in owning Apple hardware, I *do* think that Thunderbolt has its merit, and it would be nice to have it more widespread, even if every body is going to use USB3 instead.

But you need either to add expensive multiplex chips, or you need the graphic card manufacturer to start exposing the underneath PCI-e on their display connectors.The latter is the better long term solution, but it will take some time until everybody starts doing it and the components reach market.

---

[*] Well, I know but that's how it's currently perceived.

Re:Same argument for USB3

[dbIII]

Two years and that list is very short. I was hoping for a pile of monitors, a laptop at less than egotist showoff price and at least one "laplink" cable.
I was wishing for something like thunderbolt instead of a crappy USB video dongle when someone wanted several screens with different resolutions on their laptop - stuff like the Matrox external things don't do dissimilar multiple monitors.

A bunch Thunderbolt boards

[Anonymous Coward]

All Intel socket though.

Asus Maximus V Extreme
Asus P8Z77-V Premium
Asus P8Z77-V Pro/TB

  Gigabyte GA-Z77X-UP5-TH
  Gigabyte GA-Z77MX-D3H-TH
  Gigabyte GA-Z77X-UP4-TH
  MSI Z77A-G45 Thunderbolt
  MSI Z77A-GD80
  Intel D33217CK
  Intel DZ77RE-75K

  ASRock Z77 Extreme6/TB4

Re:A bunch Thunderbolt boards

[Psyborgue]

I have the Intel DZ77RE-75K. Great board, but I must say I haven't used the thunderbolt connection (yet). Perhaps for an external raid enclosure in the future, but for the moment, the internal raid is working just fine.

Bad choice of word

[dbIII]

Swap "drive speed" for "SATA" - I meant to indicate that the same drive plugged into SATA versus a USB3 enclosure or converter performs far better (greater speed and more continuous activity). That suggests that at least currently USB3 is the bottleneck by an obvious margin. Your anecdotes may vary, and if they do could you please tell me what hardware works as promised.

Re:Bad choice of word

[Anonymous Coward]

He didn't say USB3, what is so wrong with eSATA that it needs to be replaced with an expensive technology for no appreciable difference in performance. Thunderbolt doesn't have any advantages for the scenario described where you plug in a single drive to copy movies etc, in fact if you bought thunderbolt expecting that to give a huge performance increase you would be very disappointed.

Re:Bad choice of word

[bloodhawk]

SATA often appears quicker as external drives, especially USB are usually low speed drives for less power consumption as they are usually driven of the limited power of a USB port. in practise in equal comparisons there is very little difference between SATA and USB for speed. both are limited by the physical performance of the drives rather than the technology itself. eSATA can do up to about 300MB/s from memory, while USB3 can theoretically reach 600MB/s but has the overhead of conversion, in reality they are both well and truly limited by drive performance rather than technology of the connection speed.

Re:Bad choice of word

[dbIII]

Is that from reading something about the design goals or from observation? I've used a couple of mains powered USB3 to SATA caddies that take several hours longer with clonezilla than popping the case and plugging a SATA cable directly into a drive to do the same image with the same model of drive. If it's from observation please let me know the name of some USB hardware that allows the disk to be the bottleneck, because I've never seen anything that comes close.

Re:Bad choice of word

[Anonymous Coward]

That would suggest you had some really low quality USB3 chips/connectors in your machine as USB3 has to go through a conversion process which is dependent on your hardware, USB3 is also more resource intensive so machine quality is far more relevant for it. if you buy garbage then you get garbage speed

Re:Bad choice of word

[bloodhawk]

Well that sounds more like a problem with your computer than USB3 itself, USB3 is dependent on the quality of your chips and is far more resource intensive than SATA. I am currently copying ISO images off my Western Digital passport USB3 drive, I am getting just over 80MB/s. Just for shits and giggles I tested it plugged directly into SATA and it is giving me 85MB/s. The absolute peak transfer rate the drive is rated for is 135MB/s (drives are slower at various points in each platter so realistally not achievable). there is no cable/connector you could plug into this drive to make this many times faster, it is technically not possible, you could maybe gain some small fractional boost by a protocol with less overhead but barring upgrading the physical drive itself there is simply no way.

Re:Bad choice of word

[dbIII]

Fair enough, but I'm getting speeds lower than that with new 7200rpm desktop SATA drives that sustain around 100MB/s for a few hours via SATA. Most motherboards I've used are the same brand (ASUS) and within a few months of each other so probably have the same USB3 hardware. It's been a disappointment so far.

Re:Bad choice of word

[Anonymous Coward]

It could also be the quality of the external cradle your using, it has to bridge from USB to SATA. USB has a lot more components involved which can interfere with performance, but if you are getting such a massive difference between USB3 and SATA then you definitely have something either in the cradle or the computer that aint up to scratch.

Re:Bad choice of word

[dbIII]

I'm blaming the USB3 chipset on the motherboards since it's not just the caddies that are slow - USB3 rated external drives are also pathetic with the machines around here.

Multi-monitor: use it already.

[DrYak]

I was wishing for something like thunderbolt instead of a crappy USB video dongle when someone wanted several screens with different resolutions on their laptop - stuff like the Matrox external things don't do dissimilar multiple monitors.

Don't use the Matrox thing: it takes an already made video signal from the laptop and spreads it over several monitors. (So if resolution do not align, you're toast) you're also still using the limited graphics power of your laptop.

Thunderbold isn't the only technology exposing PIC-e lanes. Express Card can (among other, like USB2) expose PCIe too. (Juste like 32/bit PC card were PCI, and 16-bit PC cards where ATA).
You can find nice Expresscard-to-PCIe-slot adapters.

Plug a nice mid-to-high range graphic card (preferrably same GPU maker, and more or less the same generation, so that it plays nicely with the card inside your laptop, and works correctly in multi GPU setup. Check with your GPU provider which card can easily be mixed in multi GPU configs) and voila: you have all the multi-monitor setup you want to have, with all the crazy GPU processing powers you can buy.

Re:Multi-monitor: use it already.

[drinkypoo]

Have you actually looked into the ExpressCard-to-PCIe adapters? They are exorbitantly priced for something which is doing almost nothing, with the cheapest at $100. ExpressCard doesn't provide PCIe x16, so there is no point whatsoever installing a mid-to-high range graphics card. You cannot feed it over the ExpressCard socket, period, the end. What you want is a PCIe x1 graphics card; some few PCIe x16 cards can be hacked into one much more cheaply than just going out and buying one, but it's nontrivial. The ExpressCard to PCIe x16 sidecars are more expensive than just buying a new laptop with a real GPU, so that's utterly uninteresting.

Re:Multi-monitor: use it already.

[dbIII]

No because the machine I was setting up was a stupidly thin thing that did not have an expresscard slot, or the older slot that could take a Villiage laptop graphics card already on the premises gathering dust but perfectly capable of everything short of recent 3D games at 2048x1280. You are correct that it is worth getting something already expandable - so yes I suppose I was wishing for either thunderbolt, USB that doesn't suck, expresscard slot, or even the old PC missing in action card slot.

Re:Multi-monitor: use it already.

[Anonymous Coward]

Have you actually looked into the ExpressCard-to-PCIe adapters? They are exorbitantly priced for something which is doing almost nothing, with the cheapest at $100.

That's because almost nobody wants them. Things can only get cheap when they're made in quantity. If there was a large market for these, they'd be very cheap.

ExpressCard doesn't provide PCIe x16, so there is no point whatsoever installing a mid-to-high range graphics card. You cannot feed it over the ExpressCard socket, period, the end. What you want is a PCIe x1 graphics card; some few PCIe x16 cards can be hacked into one much more cheaply than just going out and buying one, but it's nontrivial.

You're a little confused about what is possible. I have a copy of the PCIe spec. All devices have to support link width negotiation, and all devices wider than x1 must (no ifs, ands, or buts) be capable of negotiating x1 link width as well as their native width. Support for intermediate widths is optional, but GPU chips almost always have that feature. Otherwise, they'd be forced to drop down to x1 when installed in SLI motherboards that provide x4 or x8 electrical link width in physical x16 slots.

Providing less electrical lanes than the physical connector supports is explicitly legal. Basically the only requirement is that lane 0 (the one that's on the x1 connector) has to be there. During initialization, each lane independently detects link status and trains itself. When the lanes are up and running, width negotiation merges them into groups larger than x1 (where possible, and if desired by both ends).

To build an ExpressCard-to-x16-GPU box, all you need is connectors and passive wiring. No active circuitry except for a power supply. Very cheap in principle. Expensive in practice, only because next to nobody buys them.

The hacks you speak of are people cutting x16 cards so the edge connectors fit in x1 slots. The reason it doesn't always work is that cutting through a multilayer PCB is risky. If the layout engineer ran any important connections through the piece you're cutting out, you're done. If you short power/ground planes together, or to other signals (a real risk when cutting and filing), you're done.

Running a x16 card at x1 doesn't hurt graphics performance as much as you might think. Most PCIe traffic is used to upload data to the card's local RAM. In high end cards, the GPU can access local RAM at more than 100 GB/s, but even x16 PCIe 3.0 is only capable of about 16 GB/s. This means developers put a lot of effort into optimizing away data movement, carefully controlling data set size to keep it nearly all local to the card. So long as you're running software well-optimized for the GPU, you're not going to lose as much performance as you might think. When a few review sites tried experimenting with this, high end video cards forced to x1 ran games at something like 75% of the performance of x16. (Paradoxically, it also means you'll tend to lose more performance potential on cheaper cards that have less RAM.)

Re:next generation of graphics cards will want pci

[drinkypoo]

I actually wouldn't be entirely surprised if the oil-immersion GPU happens eventually, though.

There are many reasons why it probably will never happen in a mass-market device, but weight and leaks are the two big ones.

There's really no need to immerse the whole card, nor any practical benefit given that the PCB blocks the rear of the GPU from receiving cooling from that side in any case. It's cheaper and easier to build a more parallel heat pipe cooler, and just as effective if not moreso.

As well, external GPUs are only a thing when they need to be really, really big. If we the consumer actually ever wind up with massive GPUs and tiny CPUs (improbable but possible) then we'll just be buying motherboards with GPUs on them, and plugging in CPUs or possibly CPU cards with memory on.

That still has value

[SuperKendall]

I wouldn't suggest it. It'll only take two SSDs to saturate a Thunderbolt bus (or 4 SSDs with 20 Gbit Thunderbolt).

Sure for an enterprise that is not very useful. But for a small business or semi-pro consumer, a 2SSD RAID operating at full speed of a 10Gbit bus is a real step up from eSata or USB 3.0...

There are plenty of valid uses for such systems for photographers and videographers.

Re:A video card can max out the bus and be underpo

[Anonymous Coward]

You seem to be contradicting yourself. Is x4 faster than x16?

What does the mobile GPU have anything to do with how fast data is transferred? (Unless it has no integrated memory)

Same mistake they made before!

[Anonymous Coward]

They would not let people run their precious MAC OS on anything but Apple hardware and it nearly drove them out of business.

You ever cracked the lid, or a hinge on a laptop?

[Anonymous Coward]

I haven't, because mines made of metal.

Cheers

Re:A video card can max out the bus and be underpo

[smash]

You misunderstand my point. PCIe x4 is slower than x16. I'm saying that even when running over a much slower bus than an internal x16 slot, the GPU will still be far faster in actual use than any mobile GPU chipset (i.e., the reason you'd want to run an external GPU over thunderbolt - you have a portable machine that has no, or slow discrete GPU), because the PCIe bus transfers are infrequent as everything is cached in VRAM.

ExpressCard-to-PCIe adapters

[DrYak]

Have you actually looked into the ExpressCard-to-PCIe adapters? They are exorbitantly priced for something which is doing almost nothing, with the cheapest at $100.

Yup. I've looked.
They probably cost a fraction of that if you can source them from asia.
ebay is your friend (Note: that last one taps into mini-PCIe instead of express card).

Once you factor in the price of the card you're throwing in, a sufficent powersupply to power it (a modern GFX card can eat up to hundreds of W), and a nice pimped enclosure, it's not that expensive.

ExpressCard doesn't provide PCIe x16, so there is no point whatsoever installing a mid-to-high range graphics card.

Points for using a mid-to-high range:
- the whole discussion is about driving multiple hires monitors. You need connectors for that. mid-to-high range GFX card have plenty of connecting options. entry-level card might have 2 hdmi connectors, if you're lucky.
- if gaming is desirable, you'll need a card with enough pixelrate to keep painting said multiple monitors at a decent frame rate.
- as explain by other: PCIe 16x only really matter at loading time, and more memory is always desirable.
- for desktop use, only, PCIe 1x is enough. And (provided that it has enough connection options) a mid-range card is enough too.

What you want is a PCIe x1 graphics card; some few PCIe x16 cards can be hacked into one much more cheaply than just going out and buying one, but it's nontrivial.

Non trivial as in:
- sticking it inside a connector which is already mechanically compatible with PCIe 16x (although only electrically connected to PCIe 1x) ?
- stickinig it inside a connector which is mechiannically PCIe 1x, but has a hole on the side so you can stick wider cad (although only the first PCIe 1x part of the card will be both mechanically and electrically connected) ?

PCIe technology was designed on purpose to be trivial to setup. The only drawback is that the card will only work at PCIe 1x speeds. But as said before: for desktop use, it's sufficient.
(unless you want to run CUDA/OpenCL on it).

The ExpressCard to PCIe x16 sidecars are more expensive than just buying a new laptop with a real GPU, so that's utterly uninteresting.

Please, please. Show me a laptop which:
- costs around 100$
- has a bat-shit crazy powerful GPU that can drive 3x 24"/300DPI screens
- has the necessary connector for the monitors (3x HDMI or 3x Display Port ++)
- has a replacable/upgradable GPU
- is relatively thin

ieee1394 flash drives?

[girlinatrainingbra]

hmm, I wonder if they ever had any "Firewire"/ieee1394 Flash Drives? They would be useful for booting up older macs that had firewire but do not support USB boot disk volumes.