USB 3 in 2008, 10 Times as Fast

psychicsword writes "Intel and others plan to release a new version of the ubiquitous Universal Serial Bus technology in the first half of 2008, a revamp the chipmaker said will make data transfer rates more than 10 times as fast by adding fiber-optic links alongside the traditional copper wires." "The current USB 2.0 version has a top data-transfer rate of 480 megabits per second, so a tenfold increase would be 4.8 gigabits per second." This should make USB hard drives easier and faster to use."

Comments

Great.

[TaoPhoenix]

Incredible and grand news if it's on target, and doesn't dissolve into vapor.

Cue the Media Copying Discussions.

(Someone fast on their math: How long would that take to copy a new 0.90 Terabyte drive?)

Re:Great.

[jimstapleton]

1.5 seconds if you all of your components were fast enough. The drive won't be.

Re:Great.

[morgan_greywolf]

1.5 seconds if you all of your components were fast enough. The drive won't be. Exactly. There's still the slowdown associated with the mechanical aspects of the hard drive -- spin rate (RPM), average seek time (ms), etc. On top of that, most hard drive controllers are limited by the technology they use. For instance, a SATA hard drive, even plugged into a USB 2 or 3 port, is limited to 150 MB/s -- but, that's burst speed, not sustained transfer rate.

Re:Great.

[Technician]

Cue the Media Copying Discussions.

The RIAA and MPAA just joined Steve Balmer in needing new office furniture.

Re:Great.

[Anonymous Brave Guy]

On top of that, most hard drive controllers are limited by the technology they use. For instance, a SATA hard drive, even plugged into a USB 2 or 3 port, is limited to 150 MB/s -- but, that's burst speed, not sustained transfer rate.

Indeed. And realistically, it's going to be a pretty short burst: most hard drives today only have something like 8–16MB of cache that might be filled by a smart lookahead algorithm, so your best case with current hard drive technology is that you'll get perhaps 1/10 of a second of high-speed data transfer before hitting the physical barriers.

I'm not sure this is directly applicable to this discussion, though, because AFAIK all current USB drives use different storage technology anyway. It's going to be the limits of that technology that tell us whether USB3's theoretical speeds will actually be useful with storage hardware available in the same time frame.

Re:Great.

[walt-sjc]

USB attached arrays (like drobo.)

USB Ramdisks.

Frankly, this is just "USB fiberchannel". Why not USE fiberchannel??? Surely in mass "consumer" production we can get the chipset / transceiver / cable cost down... It would be nice to come up with better connector technology that protects the optics better however, but LC isn't THAT bad, and can be had for around $16 for a 3M long cable.

Re:Great.

[BRUTICUS]

Is it me or isn't 4.8 gigs a second faster than how fast your computer actually reads info off of your hard drive? Why not just have flash harddrives using USB instead of IDE?

Re:Great.

[oliverthered]

because flash is really slow.

Re:Great.

[BRUTICUS]

as far as I know my computer can't read 4.8 gigs a second. So why not use USB interface instead of IDE? I don't get it.

Re:Great.

[thegnu]

It would be nice to come up with better connector technology that protects the optics better however, but LC isn't THAT bad, and can be had for around $16 for a 3M long cable.

If the price of USB cables is any indication, make that $70 at your local electronics retail outlet. :)

Seriously, how many non-techies do you know that think it ACTUALLY costs $18 for a 6ft USB cable?

Re:Great.

[BosstonesOwn]

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

No Flash is slow to write , very fast to read. Hence Windows use of it for "ReadyBoost" caching. There is extremely low latency just not enough bandwidth to sustain high levels of I/O.

On the other hand , by introducing fiber into the link doesn't that take away the greatest part of usb ? being able to just fold up the cable and stuff it in your pocket along with a small hard drive ? I know I use it for restoring machines after catastrophic failures (yeah windows) and some times I don't go right back to my desk with the cable and drive and have to toss it in my pocket. I can't do that with fiber, it would fracture.

Re:Great.

[Corporate Troll]

I really should buy stock in IKEA....

Re:Great.

[antime]

The quoted speed of USB 3 is probably the bus speed, i.e. it's shared by all devices connected to the same host. So one disk won't saturate the bus, but if you plug in a bunch of them the bandwidth won't seem so incredibly massive anymore. Then you have to consider the bandwidth reserved by isochronous devices etc.

Re:Great.

[pipatron]

Uh, what? 0.9TB/(4.8Gb/s) = 25.6 minutes, not 1.5 seconds, even IF the platters would cope with the new speed.

Re:Great.

[Anonymous Coward]

32 Minutes.

Re:Great.

[jonnythan]

15 seconds for those of us who can do basic math ;)

Re:Great.

[letxa2000]

This should make USB hard drives easier and faster to use."

I honestly don't see where a fiber-optic cable is going to make a USB hard drive easier to use, but the extra speed sounds nice.

Re:Great.

[Technician]

I really should buy stock in IKEA....

I doubt Steve and the heads of the MPAA and RIAA shop at a mass market store like IKEA.

You might look at places that have furniture like this;
http://www.drexelheritage.com/index2.asp

instead of this;
http://www.ikea.com/

If money was no object, which chair in the pictures would you rather sit on?

Re:Great.

[Anonymous Coward]

Great, but can the computer handle that speed ?
Yep fiber and that part is just fine , but now the issue may becomes the ability of the computer to buffer it
Obviously a hardware memory buffer can be placed in the USB hardware itself , but time will tell
4.8 gigabits per second is fast
thats 2.08e-10 or .208 Nanoseconds per bit
I wonder if people realize how fast that is and that it's faster than the CPU clock itself ?
  the latency Pc hardware can't keep up with that .
Nor can the CPU th clock rate a ! GHz CPU is 1 nanosecond
the data rate of this stuff is .2 nanoseconds Obviously the CPU cant receive it as quickly as hte data arrives so it must be buffered and that means it's going to be slow
, Let's see it work in practice, the Pc may be somewhat slower , so the actual in practice actal Throughput is the issue and may be much slower than the data rate
Also buffering this data costs money the USB device must have lots of fast memory built in
to ralize it

Re:Great.

[BRUTICUS]

thanks for explaining that.

how about CD-Roms and DVD players?

I use DVDs to back up information now, but if they can make a device that is cheaper or the same price as a DVD but plugs into a USB drive why bother with CD / DVDs?
I'd prefer to buy our games and music on USB and play it from there. I bet USB doesn't suck as much battery power as a DVD-rom does either...increasing battery life for laptops.

Re:Great.

[Corporate Troll]

True, but I understand it's for throwing around... In that case IKEA is most likely to be more cost-effective ;-)

Re:Great.

[JasterBobaMereel]

Err the PCI-X bus is only rated at 1.06 GB/s (or 4.3 GB/s for PCI-X 2.0) so what exactly is the point?

Re:Great.

[notasheep]

Quick math check: .9TB = 900GB
900GB/4.8GB/s = 187.5 seconds for the transfer
187.5/60 s/minute = 3.125 minutes

Right?

Re:Great.

[Retric]

Almost but 4.8 gigabits/s = .6gigabytes/s .9TB = 900 gigabytes
900 gigabytes / (.6 gigabytes/s) = 1500s or 25min.

Re:Great.

[mikael]

You can get an USB to IDE adapter in order to convert a 2.5" IDE laptop drive into an external hard drive. It just clips over the ID adapter and plugs into a USB cable.

Re:Great.

[pthor1231]

That and it's probably more likely to break coming from IKEA than a quality furniture place :)

Re:Great.

[jonnythan]

Yeah, this one is correct. I divided by 480Gb/s instead of 4.8Gb/s because I glanced at the summary and saw "480."

It's not 15 seconds, it's 1500 seconds.

So much for being able to do basic math.

GP forgot that a byte is 8 bits.

Re:Great.

[dintech]

Maybe drives are a bad use case for this tech then. However, it would be great for HD audio and video transfers between devices.

Re:Great.

[Agripa]

Err the PCI-X bus is only rated at 1.06 GB/s (or 4.3 GB/s for PCI-X 2.0) so what exactly is the point?

Are you confusing PCI-X (PCI Extended) with PCIe (PCI Express)?

PCI-X is typically operates from 4.3 Gb/s (64 bits at 66 MHz) to 8.5 Gb/s (64 bits at 133 MHz) and is normally only used in servers or workstations.

PCIe starts out at 2 Gb/s (twice the speed of PCI) and easily expands to handle higher transfer rates by using multiple links. x4 and x16 links are common.

Re:Great.

[DaveWick79]

Flash may be slow to write, but it is still volumes faster than a hard disk write.

Re:Great.

[xenn]

Yeah, 1.06 GB/s ought to be enough for anyone!

Re:Great.

[JasterBobaMereel]

The most number of lanes supported by PCIe is x32, so 250MB/s x 32 x 2 (bi-directionality) is 16GB/s for a theoretical maximum transfer rate. - Please note this is theoretical maximum ...

PCI-X maximum is as I said 4.3GB/s ...

But the real killer is maximum implemented storage transfer speed bus is Fibre Channel 4GFC (4.25 GHz) - 425 MB/s

Re:Great.

[masticina]

Lets not forget ONE external hd on itself might mere doe 80Mbyte per second..still it is better then he 20Mb most CRAPPY USB-IDE controllers can handle. They definitly can improve on that part! USB definitly is not the final limit for the external hd, it is the low quality crappy controller inbuild in most boxes!

Re:Great.

[empaler]

First completely logical post in the thread. Thank you :)

Re:Great.

[BronsCon]

And computers are never going to get faster? Well... I guess Moore's law has failed already then.

On the other hand, the CPU does not process each BIT as it enters the system. They're sent to a buffer on the device, then to RAM, then processed as a datatype, not individual bits. So, at 1 tick per 1 nanosecond, 1 bit per 0.2 nanosecond (that's 5 bits per nanosecond)... you're looking at less than 1 byte per tick. I would assume it's to be processed in 64 bit chunks on a modern 64bit CPU. That means it will be processed only when at least 64 bits are available to process, so, every 12.8 nanoseconds.

That's pretty reasonable. Also, I'm guessing, if you're doing something that's maxing out that bus, you probably have multiple cores running at over 2Ghz each and enough RAM to handle buffering. What would happen with a single 1Ghz core is rather irrelevant; even the low-end laptop I picked up in February had dual 1.7Ghz cores.

Re:Great.

[nxtw]

USB 2 is 480 megabits/sec max; SATA 150 is 150 megabytes/sec max; SATA 300 is 300 megabytes/sec max. A single drive will easily saturate USB 2 or FireWire 400 and some newer drives can saturate FireWire 800.

The SATA bus can carry data from more than one drive; port multipliers will connect up to 5 drives directly, or a SATA device could be some sort of hardware RAID implementation that presents itself as a single drive to the host. (My maximum throughput with SATA has been 120 megabytes/sec on a supposedly SATA 300 link with a PCI Express x1 host adatper, with 5 drives that easily sustain 60 megabytes/sec each.)

So, USB 3, at 4.8 gigabits/sec, can be saturated easily with today's SATA devices, even if it takes more than a few drives.

Re:Great.

[cnettel]

In burst (no seeking), not in sequential writing, unless you put multiple blocks in massively parallel configs, something that's generally not done in current consumer SSDs.

Re:Great.

[encoderer]

Volumes Faster? nice pun...

Re:Great.

[poolmeister]

You PC with one or two hard drives may have trouble getting to 4.8Gbps transfer rates but how about RAIDed disks?
Current SCSI over FC SAN adapters only go upto 4Gbps so maybe this could be the beginning of USB SANs?

Re:Great.

[Anonymous Coward]

The original poster is correct
A 1 GHz CPU or N GHz CPU required 2- 10 or more CPU cycles per operation
  Faster data must be putstored somewhere in real time , you seem to understand that , but missed entirely how that happens. you obviously don't understand the concept well enough .

It must slow down the overall process if the CPU cannot keep up with the data rate period , this is called latency , it has nothing to do with Moore's law
And No
Each bit is processed exactly 1 bit at a time in USB as there is only 1 physical line USB is a serial communications device and so some type of serial to parallel conversion therefore must take place to re-assemble data bits into into bytes and words What we get as a speed increase is much more complicated than that .
The original poster is absolutely correct
If the CPU is 1 GHZ it can NEVER do anything faster
  in fact there are 2 to 10 CPU clock cycles required per data transfer in any type of CPU and another initial 4 - 20 to process the interrupt that initiates the data transfer
You need to sit down with a good computer book and consult with someone knowledgeable of computer hardware at the bit level.May I suggest a title like:
  Microprocessors hardware and interfacing and a good tutorial on serial communications
Best regards

Re:Great.

[Anonymous Coward]

On the other hand , by introducing fiber into the link doesn't that take away the greatest part of usb ? being able to just fold up the cable and stuff it in your pocket along with a small hard drive ?

One of my biggest gripes with USB (besides all the technical blunders) is that there's no "the cable". My iPod has a USB-to-iPod cable. My camera has a USB-to-camera cable. My Visor (back when I used it) had a USB-to-Visor cable.

WTF, guys? The point of a standard cable is so I only need to carry one fucking cable.

Re:Great.

[Fifty Points]

Have you ever handled fiber? It's not that fragile.

Re:Great.

[pipatron]

Have you even read the book you recommend? Ever heard of things like DMA? Bus controllers? Did you know that a CPU can actually process more than one bit during one cycle? Yes, a CPU can never handle more than it can handle. Logic. Even a slow CPU by todays standard can shuffle around data faster than the 480MB/s that this future technology can stream. Or do you think that the CPU reads 4 bits at a time from my gigabit NIC, 250 million times/second? Or one bit at a time from my 3GHz SATA-disks?

Re:Great.

[Jeremy Erwin]

I know that when I use a USB drive, I first unplug my mouse and keyboard, so that the drive gets to use the full bandwidth. Wouldn't want to waste even the slightest bit of speed.

I also close other programs before opening new ones.

Re:Great.

[Grishnakh]

Not only that, but for the devices that do use a standard cable (like printers and scanners), the cables are obscenely expensive, at least if you buy them in a retail store.

As for your other examples, there some (sorta) good reasons for all the different cables. For the iPod, they wanted to have a single standard dock connector, so you could connect your iPod to your computer, your stereo, etc. all through one standardized connector. So they just made one version for USB-to-iPod. For the camera, they probably wanted to use a smaller USB connector on the camera than the standard (large) "B" connector, so they probably used a "mini-B" connector. The problem with the latter is that, while there are standardized USB mini connectors, there's some variation and not everyone uses them. I think my camera uses some non-standard mini-USB connector for instance, while my iRiver uses a standard mini-B.

Intel really didn't do a very good job with the USB specification if you ask me.

Re:Great.

[Bassman59]

1.5 seconds if you all of your components were fast enough. The drive won't be. Exactly. There's still the slowdown associated with the mechanical aspects of the hard drive -- spin rate (RPM), average seek time (ms), etc. On top of that, most hard drive controllers are limited by the technology they use. For instance, a SATA hard drive, even plugged into a USB 2 or 3 port, is limited to 150 MB/s -- but, that's burst speed, not sustained transfer rate.

you people are all missing the point.

It's obvious that the 4.8 Gbps link is faster than the device ... but recall that all USB devices on a port share the bandwidth. A faster link will allow a lot more devices to simultaneously transfer data at their maximum-possible speed.

One example: You'll be able to put a multichannel audio I/O device and hard disks on the same bus without worrying about dropouts, etc.

As for "Isn't 4.8 Gbps faster than the computer can handle?" That's true, and it's already the case when you have Gigabit Ethernet plugged into 33 MHz/32-bit PCI. But remember that 4.8 GBps is the wire speed, not the data rate, which will ultimately be controlled by how fast the host computer can get data onto and off of the bus.

Re:Great.

[MetalPhalanx]

But if everything worked off of "one cable" then they couldn't charge you for their own semi-proprietary version. My Motorola phone uses the a miniature version of the regular USB device end, and they want to charge some ridiculous price for a cable (which is of course only available through them). Luckily for me, my POS Sony camera has a cable which fits.

Re:Great.

[Anonymous Coward]

Good try, but...

- Apple could have physically put a USB jack right next to a custom jack for the other stuff. Then you could both use a plain USB cable, and could also dock it in your stereo and get sound from it. Their ADC connector (= DVI + USB + power) had the same issue, so finally they gave up and went with DVI + USB + power, but with cables that had all 3 together.

- My camera does use a mini-B, but, except on the smallest cameras (which mine isn't), there's plenty of room for a standard-B. That's just a lame excuse. (My original iPod had a full-size Firewire jack, and my camera isn't cramming things in nearly as tight as an iPod.)

- I see you didn't even try to come up with a lame excuse for Handspring. Yeah, they were beyond hope. :-)

For even more fun, try to explain APC's "USB" cables. They're USB-A-to-RJ-45. Yes, you read that right.

Intel really didn't do a very good job with the USB specification if you ask me.

Now, now, there are plenty of companies listed on the spec, and I'm sure they're all morons -- Microsoft, NEC, HP, the whole bunch. I like to at least think it was a software company like Microsoft who wrote the part about devices being required to report their max-power in milliamps. *forehead slap*

Re:Great.

[Wdomburg]

No, he had it right. The initial speed of PCI-X (which debuted at 133MHz, not 66MHz) is 1064MB/sec (note: bytes, not bits) and the 2.0 spec does cap at 4.3GB/sec.

Where he seems confused is that the speed of USB is in megabits. A tenfold increase would bring it to 600MB/sec which would comfortable served by PCI-X 1.0, PCI-e 1.0 4x or PCI-e 2.0 2x. He also fails to take into account that USB is generally integrated directly into the chipset rather than being hung off the PCI* bus, so it's pretty irrelevent to the discussion.

Still, here's a run down of the various speeds offered by PCI, PCI-X and PCI-e. The 5Gb/sec variant of PCI-e isn't on the market yet but Intel, AMD and Nvidia have chipsets scheduled for Q4.

PCI, 32-bit, 33MHz - 133MB/sec
PCI, 64-bit, 33MHz - 266MB/sec
PCI, 32-bit, 66MHz - 266MB/sec
PCI, 64-bit, 66MHz - 533MB/sec

PCI-X, 64-bit, 133MHz - 1064MB/sec
PCI-X, 64-bit, 266MHz - 2.15GB/sec
PCI-X, 64-bit, 533MHz - 4.3GB/sec

PCI-e, 2.5Gb/sec, 1x - 250MB/sec
PCI-e, 5.0Gb/sec, 1x - 500MB/sec

Re:Great.

[aj50]

I'm guessing it's not the same for all flash, for instance this usb memory stick is rated at speeds of 900KB/s for writing and 4.3MB/s for reading, much slower than a hard-disk (at least for reading, not sure about hard-drive write speeds). Other memory sticks on that site list read speeds below what I'd expect from a hard drive.

Re:Great.

[Grishnakh]

Good try, but...

- Apple could have physically put a USB jack right next to a custom jack for the other stuff. Then you could both use a plain USB cable, and could also dock it in your stereo and get sound from it. Their ADC connector (= DVI + USB + power) had the same issue, so finally they gave up and went with DVI + USB + power, but with cables that had all 3 together.

Hey, I'm not trying to say these were great reasons for the cable mess, just reasons.

You're right, a dock connector plus a plain USB jack would be ideal. But then, that would add extra cost, and take up more room in an already tight design. The original iPod with a full-size Firewire jack was thicker than today's ultra-thin models.

However, personally, I use an iRiver H330 (was an H320 but I upgraded the HD), and it has a standard mini-B connector on the bottom, and a standard phono plug for power. There's an available docking station (I have one) which has the male versions of both these connectors, so it is possible to not even have a special docking connector, and just use standard cables. Unfortunately, this doesn't allow for connecting to a stereo, as the line in and line out jacks are on the top of the device, as is the jack for the optional remote.

The iPod would have been better with a separate mini-B jack in addition to the dock connector, but as I said, this adds cost (10 cents per 100 million iPods = $10 million). Plus, and probably more importantly, you wouldn't need to buy a special cable from Apple for $$$ when you break yours.

- My camera does use a mini-B, but, except on the smallest cameras (which mine isn't), there's plenty of room for a standard-B. That's just a lame excuse.

Most of the cameras I see these days are extremely small and thin: about 3" wide, and 1/2 - 3/4" thick. I imagine space inside those things must be very tight, so I think that trying to cram a standard-B in there is out of the question. Standard-B jacks are really huge; much thicker than a standard-A, which is for more comparable in size to the Firewire jack you mention.

For even more fun, try to explain APC's "USB" cables. They're USB-A-to-RJ-45. Yes, you read that right.

Hahaha. Sounds like APC just decided to make nonstandard cables using totally standard (read: cheap) off-the-shelf parts, just so they could jack you on replacements.

Now, now, there are plenty of companies listed on the spec, and I'm sure they're all morons -- Microsoft, NEC, HP, the whole bunch. I like to at least think it was a software company like Microsoft who wrote the part about devices being required to report their max-power in milliamps. *forehead slap*

Well, you know what they say about intelligence and committees: take a bunch of geniuses, put them together in a committee, and their combined intelligence is about the same as someone in an institution for the mentally impaired. The sum is much lower than the individual parts.

From what I remember, though, Intel did the bulk of the work on USB, as they wanted a high-speed hot-plug serial bus to replace the clunky old serial and parallel connectors, but in the process tried to make it use extra CPU horsepower to spur people to buy faster CPUs.

Re:Great.

[Wdomburg]

There are some 10GFC products on the market, just not HBAs. I believe Emulex and Qlogic will be shipping their 8GFC HBAs this quarter as well.

Re:Great.

[PitaBred]

I don't have a fibrechannel connector on my machine right now. And USB3 is backwards compatible, so I could either get a USB3 drive and know it'll work on my USB2 ports, or get a USB3 based machine and know my current USB2 devices will still function. Basically, USB3 is a lot more useful than fibrechannel because it's general-purpose and ubiquitous.

Re:Great.

[Obsi]

1995 called, it wants its solid state game storage medium back.

Re:Great.

[Anonymous Coward]

Physical limitations aside -- does no one do math anymore? .9TB ~= 900 GB. 4.8Gb/s ~= 600MB/s. Try 1500s.

Re:Great.

[Anonymous Coward]

There is a minimum coil size listed on the packaging for a reason. If you coil it too tight or fold, it will fracture.

Re:Great.

[Propaganda13]

Sure you can get a USB device that is the same price as a DVD recorder, but there's different ideas behind each device. DVD recorders have "unlimited" storage space and that storage doesn't have to be in one spot. I can make backups everyday and not run out of space and I can keep those backups in separate locations. I have no problem burning a 1KB file to a DVD-R and giving it to someone. I'm not going to do that with a USB flash device.

Re:Great.

[Hucko]

Grishnakh's committee law: "The sum of the parts is inversely proportional to the individual parts"

Re:Great.

[toddestan]

Is it the Mini-B connector like this? I'm assuming so since every Sony camera I have run accross uses that kind of plug. If so, that's a standard connector that's used by a lot of things (I have a 2 camera's, a MP3 player, a hub, and a card reader that all can use the same cable).

Re:Great.

[BronsCon]

Ahh... someone else who understands >1bit CPUs and modern system archetechtures.

Glad to have you on the team.

Re:Great.

[Hucko]

I'm more worried that people believe products need those prices to make a reasonable profit. My brother, a financial adviser, believes it. How are the Jones going to think otherwise if one purporting to understand economics can't see past the gouging?

Re:Great.

[Anonymous Coward]

For instance, a SATA hard drive, even plugged into a USB 2 or 3 port, is limited to 150 MB/s -- but, that's burst speed, not sustained transfer rate.

The Seagate 750GB drives are capable of sustained sequential transfers in the 80-90 MB/s range. So if they can drive up the bit density just a bit more, you'll be in the 150 MB/s range. It wouldn't surprise me to see consumer drives that can do sustained transfers of 200 MB/s by the end of the decade.

Re:Great.

[Hal_Porter]

I like to at least think it was a software company like Microsoft who wrote the part about devices being required to report their max-power in milliamps. *forehead slap*

http://www.beyondlogic.org/usbnutshell/usb5.htm#ConfigurationDescriptors

bMaxPower. 1 byte. Current in 2mA units. The idea is they can code from 0ma to 510ma in 2mA steps using one byte.

It would have been nice to have specified a max current of 1A and used 4ma units so USB 2.5 inch hard disks could spin up powered by one USB connector. Which they can do with Firewire.

I must say the fact that bMaxPower is in the wrong units never occured to me - I just converted the 5W figure in the Samsung data sheet to mA by dividing by 5. Actually, measuring spin up power as opposed to current seems a bit unhelpful - since it's a spike it seems more natural to talke about peak current rather than peak power.

Re:Great.

[JasterBobaMereel]

Aha!

    So we have the usual confusion between nBits - nBytes and also Mega = 1000x1000 / 1024x1024 (/1000x1024!)

    The bus speed is not irrelevant since the data still needs to go somewhere? Video card is on the bus, external device is ....

    Also the HDD is probably slower than this ?

    So where is this data going ? (and for that matter where is it coming from) at these speeds ?

Re:Great.

[Wdomburg]

Most things in modern designs are hung directly off chip set rather than sitting on a peripheral bus. E.g. the upcoming AMD 790X north bridge integrates support for 41 PCI-e 2.0 lanes, HTX, and dual gigabit ethernet. The corresponding south bridge (AMD SB700 or SB750) integrates support for up to six SATA 3Gb/s devices, 12 USB 2.0 ports and 2 USB 1.1 ports. The interconnect between the the processor and north bridge and between the north bridge and south bridge is HyperTransport 3.0. I don't believe it's know how wide the bus on that implementation is, but the specification caps at 41.6GB/s aggregate bandwidth.

Even then, there's still plenty of bandwidth on the peripheral bus. 41 PCI-e 2.0 lanes should provide about 20GB/s of raw bandwidth. Even looking at a current 4x PCI-e 1.0 video card, you've got 1GB/s of bandwidth.

Hard drives are a typically the slowest component in a system but increased densities and aggressive caches are pushing transfer speeds fairly high. There are affordable SATA drives that push 100MB/sec sustained now and high end drives are pushing more like 160MB/sec. Not enough to saturate the theoretical bandwidth of a USB3 link with a single drive, but still more than a USB2 link provides today. Not to mention that multi-disk external enclosures are becoming fairly common.

Cable?

[mwvdlee]

I though one of the major benefits of USB was that you could have everything on one type of cable so you'd just have a bunch of identical ports and it didn't really matter which was connected to the printer and which to the mouse.
Seems to me that neither the optical cable (nor the ports) will be compatible with USB 2.

Re:Cable?

[SnoopJeDi]

I can't see any reason it wouldn't be backwards compatible...USB2 just can't utilize the fiber-optic component of the USB3 wire. And surely USB3 would be smart enough to know when a USB2 wire is plugged in, and would be capped at the old transfer rate (just like plugging a USB2 device into a USB1 port)

Then again, it's all my early-morning speculation without RTFA.

Re:Cable?

[YrWrstNtmr]

The last line in the article:
"It will be backward compatible, so current USB 2.0 devices will be able to plug into USB 3.0 ports."

Re:Cable?

[Televiper2000]

There's already at least 3 different kinds of USB cables when you only consider connector types.

Re:Cable?

[Televiper2000]

It's likely established at connection. I would imagine that the Optical fiber is a separate connection much the same way that Firewire's CAT-5 and POF options are separate connections (of course their 800MBit stuff has a separate connection as well). There's still a huge benefit to keeping everything but the physical layer exactly the same. It definitely means you won't require separate USB2 and USB3 devices inside your computer.

Re:Cable?

[Stonent1]

I think the fiber part will be inside the extra space in the connector in a way that doesn't interfere with the electrical part of it. Probably when you plug in a cable the electrical part asks the connection if it is USB3 capable and if the device responds yes, it turns on the fiber transceiver.

Re:Cable?

[FinchWorld]

Ah, but it doesn't mention if USB 3.0 will work with USB 2.0/1.1/1.0, 2.0 work on 1.0, at a slower data rate mind. Suppose if 3.0 will be dependant on fibre, or just use it where available.

Re:Cable?

[Televiper2000]

I'd say it's unlikely since there's not enough room in a standard USB cable to accommodate a POF cable. That's not considering the fibre optic transceiver at either end, and the termination. It just wouldn't make sense to make it that complicated when they already have 3 different device end connections. Especially when the key advantage to expanding USB to fiber is keeping the electronics 90% the same, and the the software 99% the same.

Re:Cable?

[dwater]

does it mention if usb 3.0 will slow down if you plug a usb 2.0/1.1/1.0 on a usb 3.0 bus?

like if you plug a usb 1.X device onto a usb 2.0 bus, then everything slows to usb 1.X. IINM...

Re:Cable?

[Critical Facilities]

There's already at least 3 different kinds of USB cables
I'm glad someone else pointed this out. I think this is ridiculous, personally. The entire idea behind USB is to make it universal. Yet, my MP3 player, my printer, my camera and my TASCAM US-122 audio interface all have DIFFERENT USB connectors. So friggin' frustrating.

Re: Cable?

[Dolda2000]

Yeah, but it doesn't mention if it will be forward compatible, so that you can plug 3.0 devices into a 2.0 port, as you can with 2.0 devices and 1.0/1.1 ports.

Furthermore, I just can't help thinking it seems kind of stupid with optical links if they want it to truly take the standard over from USB 2.0. I would think it much harder to produce cheap devices that need to include optical transceivers. One of the great advantages of USB in my mind was always that it's very easy to design "client" devices -- you just hook it into your microcontroller and add a couple of capacitors and resistors. But I don't know, maybe they plan to build the optical transceivers directly into the connector, so that the device designer won't even have to think about it.

Another problem seems to be the connection between the HCI and the actual host interconnect. 4.8 Gbps would require at least two PCIe lanes. I doubt many people will want to waste a 16x PCIe port on their motherboards to connect a USB HCI card, and all too few motherboards seem to come with 4-lane connectors. Sure, that won't be a problem once motherboards start integrated USB 3.0 HCIs, but what about until then?

Re:Cable?

[DavidTC]

It's not so much the different sizes as the unneeded different sizes. I at least understand the rational for a smaller plug. A big plug for easy of use with 90% of stuff, and a tiny plug when the device itself is tiny. (Although I will point out that damn USB hubs are not tiny. By definition, they can fit the big connectors, because, duh, that's how you plug into them.)

But WTF is up with the large square connector? Is it really so important to have male connectors at both ends?(1) That was a stupid decisions to start with. People are not idiots, and they perfectly understand the idea that all cables are 'extension' cables. In fact, they probably understand that idea better than random connector changes. That's how serial cables worked for years.

Or the other smaller connector that a mp3 player of mine had? I think this is some absurd change of the tiny port from one plug to another, but I have idea why, as it's not noticeable smaller.

1) Male and female always confused me on USB, because there's male and female of the outside, which is what actually counts for the named gender, but inside that each side has something sticking out, and, what's worse, on the female plug, that little prong is what holds the male in place, from the inside. For a while I didn't grasp it was the metal shield that counted as the 'prong' for the purposes of gender, because on things like serial connectors it doesn't, the metal bits inside do.

Re:Cable?

[compro01]

the main problem is propertory connectors. there are a grand total of 4 USB connectors. A, B, mini-b, micro-b.

past that, there're 18 bazillion kinds of "USB" connectors made by stupid companies.

Re:Cable?

[Anonymous Coward]

does it mention if usb 3.0 will slow down if you plug a usb 2.0/1.1/1.0 on a usb 3.0 bus? I'm pretty sure the speed of light won't be affected by the presence of a usb 2.0 device

Re:Cable?

[ddstreet]

> like if you plug a usb 1.X device onto a usb 2.0 bus, then everything slows to usb 1.X. IINM...

This is wrong, if you plug a USB 1.1 device into a USB 2.0 "bus" then it does NOT slow everything down. Specifically there are 2 cases:

1. You plug a USB 1.1 device directly into your computer (i.e. directly into the "host controller"). In this case, the USB 2.0 host controller (technically a EHCI chip) does NOT talk to your device. Instead, the EHCI chip has one or more USB 1.1 host controller chips (technically either a OHCI or UHCI chip, and called a "companion" chip when inside a EHCI chip) and your USB 1.1 device is connected electrically to that controller. You device is not on the USB 2.0 (EHCI) bus.

2. You plug a USB 1.1 device into a USB 2.0 hub. In this case, the USB 2.0 hub creates a complete USB 1.1 environment specifically for your device. On the host-facing side of the USB 2.0 hub, all communication continues to take place at USB 2.0 (i.e. 480Mbps) speeds. When the host wants to talk to your USB 1.1 device, it uses what is called "split transactions" to talk to it. Basically (I'm simplifying), this involves sending a "start" packet to the USB 2.0 hub. Then, the USB 2.0 (EHCI) controller goes on to do other things, while the USB 2.0 hub initiates the transfer to your device at USB 1.1 speeds. And data transferred from the USB 1.1 device is stored temporarily in the USB 2.0 hub. Eventually the USB 2.0 (EHCI) host sends a "finish" packet to the USB 2.0 hub. If the USB 1.1 transation finished, the USB 2.0 hub responds successfully (either with the incoming data or a "ack" that the outgoing data was sent) which completes the transation.

(There is also a combination case of those, where the EHCI chip does not contain a "companion" USB 1.1 chip, but instead contains an internal USB 2.0 partial hub - the "transaction translator" part - that handles talking to USB 1.1 devices. For bus usage purposes, this is effectively the same as using an external USB 2.0 hub, since the USB 1.1 devices do not appear on the USB 2.0 bus.)

Re:Cable?

[Anonymous Coward]

That's it? That's your whole rant? You must not lead a very frustrating life.

Re:Cable?

[sylvandb]

the main problem is propertory connectors. there are a grand total of 4 USB connectors. A, B, mini-b, micro-b. But don't forget that each of those connectors come in male and female, making 8 USB connectors.

Now look at the variety of cable types needed...

Remember that while the host is supposed to have female A, some peripheral devices also have female A, making maleA:maleA cables needed. And people need or want extension cables, maleA:femaleA.

So while in theory three cables should be sufficient to connect all peripherals using standard connectors, reality differs.

sdb

Re:Cable?

[default luser]

The dual-bus interface would also solve one of the problems of USB2 versus 1.1 backward compatibility, where a slower device on a shared bus will slow down high-speed devices. This is because bus transactions at the lower speed inherently take longer to complete, reducing the time slices available to the high-speed devices.

With USB3, you've got fibre for the high speed, and copper for the low speed.

Re:Cable?

[Televiper2000]

Part of making USB ridiculously easy was to have a host connection and device connection since the cable does have a 5V power source built into it which could cause damage if plugged in backwards. USB doesn't actually sex their connectors. They pleasantly refer to them as "plug" and "receptacle" which is great because people were confused enough with the D-SUBs used for RS-232(1). So there's the A or flat-rectangle connector for the host and B squared tapered connector for the device. They then invented the micro-B for things like cell phones and cameras which is the small version of the B. The other small connector I've been seeing looks like the Micro-A. Which seems to mean that my USB phone is a host, or the people who made it put the wrong connector on. It's looking like I have to read the entire datasheet to figure this out. Anyone? Anyone? 1) Anecdotally. I find sourcing D-SUBs the male and female have never been mixed up. But, I've seen males referred to as "pins" and "plugs", and females referred to as "holes" and "sockets" along with the plug and receptacle lingo we see with USB. I don't even want to get into the time I thought it would be smart to get a SCSI connector for a project I was working on (literally had to print out datasheets and verify things mechanically).

Re:Cable?

[dwater]

> This is wrong,

Hrm. Interesting. I wonder where I got that information from...

Anyway, I sit down corrected.

obligatory simpsons quote

[wwmedia]

Nerd: I've developed a program that downloads porn from the interet a million times faster than normal Marge: Who would need that much porn Homer: [drools]...oohhh..1 million times faster..

I'm more concerned with latency.

[jimstapleton]

It seems current hard drives test to 40-80M/s (dunno if it's bit or byte, we'll assume byte since it is worst case for my example)., averaging between 50 and 60M/s

480Mbit per second = 60MByte per second. That can handle the average case for a modern hard drive.

4.8GBit/second - 600MByte/s? To utilize that with a drive, you'd need a RAID external enclosure!

Re:I'm more concerned with latency.

[Laebshade]

I can't find the exact article, but you should read this one about the effective USB 2.0 speed. It states that the effective maximum speed is only about 40MB/sec, and that 60MB/sec can't be achieved due to overhead/software limitations; not sure if this is true now.

Re:I'm more concerned with latency.

[W2k]

That's with today's drives. Keep in mind that this version of USB is not due to arrive on the market for another 1-2 years, and is expected to stay in use for many years after that. Furthermore, RAID enclosures aren't actually that uncommon.

Re:I'm more concerned with latency.

[Setti45]

The 480 megabits/s is the rated burst transfer speed. I can't remember now what the sustained transfer rate is supposed to be.

Re:I'm more concerned with latency.

[wrmrxxx]

The throughput of a USB connection does not equal its clock frequency as there is quite a bit of overhead in the protocol, so in reality it would be a fair bit less than the 600MByte/s approximation. Because it's a bus, the total bandwidth available can be split amongst multiple devices. With several high speed devices on the bus, 480Mbit per second might not seem so much like overkill.

The current version of USB provide connections that can operate in an isochronous mode (see http://www.beyondlogic.org/usbnutshell/usb4.htm#Isochronous) for bounded latency applications, but this wouldn't be suitable for communicating with a hard drive because it doesn't offer guaranteed delivery.

Re:I'm more concerned with latency.

[Televiper2000]

My experience with USB is that it has lousy thru-put do to the way the drivers manage data on the host side. We did some heuristic timing tests on USB to serial devices and found that most of them actually degraded the thru-put (time send and receive a packet) of an RS232 connection. We found that devices with Silicon labs ICs could get the desired thru-put, but it seemed they would take up an entire USB channel to do so.

Re:I'm more concerned with latency.

[zippthorne]

Why RAID? Why not just 10 different devices with the approximate bandwidth requirements of a single HDD?

it could be 10 separate hard disks, or some other combination of devices. Storage isn't the only thing you attach to a USB, you know.

Re:I'm more concerned with latency.

[jimstapleton]

But, 10 years ago, hard drive speeds were 20-40 MB/s I thought?

It's not like hard drive speeds are going up rapidly or anything. Although, that may change with flash based drives.

Still, I've not seen many USB based RAID enclosures. I guess part of the reason I mentioned that is that I suspect the lack of USB RAID enclosures will change.

Re:I'm more concerned with latency.

[JanneM]

How about hi-def video?

Or having more than one mass storage device, video camera and microphone, speakers and so on hanging off the same USB hub you plug into your laptop. One connector for everything, and that without the expensive, model-specific docking stations we've been using so far.

Re:I'm more concerned with latency.

[schwinn8]

Exactly... Firewire400 works well for video streaming from a DV cam because it has very little overhead. Even though USB2 supposedly does 480Mbps, it can't do DV because there's too much overhead. Bottom line is, unless USB3 gets rid of the CPU dependency and overhead issues, I won't like it. Sure, with a "ten times" the performance, this won't hinder DV, but that doesn't make it good. I hope they make it systemically-efficient, instead of just ramping clock rates to reach these speeds.

Re:I'm more concerned with latency.

[Gr8Apes]

With Firewire 400/800 (IEEE1394a/b) I got around 16MB/s with 7200 RPM SATA external drives. That's pretty much what I expected as a maximum continuous write speed from the drives. This was much faster than with my USB 2.0 ports no matter what I've tried with them.

I think any talk of USB 3.0 faster speeds for devices is pretty much vapor. Why? Because of eSata. No separate hub needed, no special hardware on drives (the primary user of high speed devices). It's serial. It's almost as fast as this proposed USB 3.0, and probably as fast given the inherent overhead of USB.

Re:I'm more concerned with latency.

[spartanhelmet]

With a number of devices working at such a fast bandwidth, i'd start to seriously worry about northbridge chipset design. As it is now, the chips are dodgy. Using a couple of raid devices, a top-end 3d card and some high speed DDR2 puts most NB chips to breaking point... and even though that's a chip closer to year old, and without any overclock, it's still ridiculous. So really... we need chip manufacturers to go back and start designing decent NB chipsets so it doesn't become any more of a bottleneck than it is. Anyone who thinks high bandwidth links like a 480Mbps connection should be put through southbridge.... should be shot.

Re:I'm more concerned with latency.

[BloodyIron]

SSD, my friend.

They're not quite there, yet.
Have you seen the charts? Very impressive sustained data transfer rates.
Just wait until they start hitting the 240gig range ;)

Re:I'm more concerned with latency.

[pipatron]

I'd like a fast graphics card on USB. And a HD video in. And two or more fast harddrives. Then add a gigabit NIC to this. Now imagine all on the same hub.

And this is what I want today, I can imagine it'll be quite much more in the Future(tm)

Re:I'm more concerned with latency.

[Joe The Dragon]

Start useing AMD cpus they have less load on them and do not get as hot. Also in a 2 or more system you can have more then one NB chip. Also putting the 3d card on the HT bus can help as well.

Any way USB needs cpu offload like fire wire or will become even a bigger joke next to Fire wire and usb at that speed. Right now firewire 400 is faster firewire 800 blows it way and the planed firewire 1600 will be even faster.

Re:I'm more concerned with latency.

[InvalidError]

I have a 17GB Maxtor drive from that era, the sustained throughput on it is closer to 12MB/s. On the 5.7GB drive I had before that, this figure was closer to 7MB/s. My current HDDs are all over the 40-70MB/s range... so sustained throughput has not increased a whole lot over the last 10 years - it roughly doubles every 4-5 years. Yes, that is quite slow progress compared to most other PC technologies.

For flash, I think I would prefer that they work on endurance before they start pushing speeds - faster flash technologies have been known to wear out considerably faster than the slower ones. The fastest flash cards are known to show reliability issues even under 1k rewrites while low-speed cards often have endurances beyond 100k. Progress in SSD performance will be in large part dictated by the cells' endurance at the faster speeds.

As far as external RAID enclosures are concerned, I would be far more interested in seeing some of these for eSATA... not the N-drives with N-links to a PCIe RAID card but one where the RAID controller resides in the external box and has a single 3Gbps/NCQ eSATA link to the host PC. Such a thing could be done with a single Virtex-5 LXT FPGA and some DRAM/SRAM.

Re:I'm more concerned with latency.

[jimstapleton]

i.e. "Our computers now have twice the power - they can now crash twice as quick"

Re:I'm more concerned with latency.

[Barny]

Mod parent up, hammer meets nail in their post :)

Bonus points if you hook a 108mb/s wireless lan adapter via USB and throw some large data files over it, watch your data speeds closely, and monitor system performance even closer.

Firewire (1394, ilink, DV port, whatever) really was the shit, not only fast, low overheads AND its a peer-2-peer setup, in a pinch you could daisy chain PCs with it for an impromptu 400mb/s lan.

Why didn't they just hang USB out to dry and get power into the eSATA spec and use that? At least then no extra chips would be needed on a mobo, external HDD would hookup with no loss in performance and we might finally see thumb drives that work natively with ANY os as... drives.

Re:I'm more concerned with latency.

[Altus]

firewire really was the shit? I have it on most of my devices. Even my cable box can output over firewire.

Re:I'm more concerned with latency.

[aadvancedGIR]

Although I totally agree, computer or braoder high tech history shows that "systemically efficient" designs usually have a hard time competing against cheaper "almost OK" ones. Plus, if it has backward compatibility with today's standard, it's already the assured next standard even if they only increase maximal speed.

Re:I'm more concerned with latency.

[Splab]

And now imagine everything going on a hub creating massive collisions....

Re:I'm more concerned with latency.

[DaveWick79]

I was thinking exactly the opposite when it comes to eSATA vs. USB3. If USB3 can handle a transfer rate far over an above what SATAII supports, and probably well beyond what SATA3 will support, I can see eSATA just dropping off the map.

System overhead on a dual core or quad core and above CPU is not going to be an issue. I'm not tremendously familiar with the SATA bus capabilities but I don't think it really has the capability of being a replacement for USB, especially considering the limits on cable length.

Re:I'm more concerned with latency.

[TheSkyIsPurple]

Even if you got those sorts of speeds, what effect would that have on your overall system speed?
If you were managing a group of devices via RAID, and able to feed them all over the bus, since USB is dependent on the main processor, how much overhead will that 10x cost you.

big deal? doubt it

Re:I'm more concerned with latency.

[Anonymous Coward]

"Hub" does not preclude the word "switched". Hint: your PC is actually just a giant hub. All your USB devices, PS/2 devices, Fireware devices, Ethernet devices, PCI devices, etc. are coming together in one giant hub. Admittedly there are a lot of sub-hubs along the way, but the point remains that it's possible to have many many extremely fast devices all communicating in one device.

Re:I'm more concerned with latency.

[dadragon]

I don't think he meant it to be a replacement for USB. USB has its place, just DV and drives aren't it. Firewire and eSATA are much better for that.

Re:I'm more concerned with latency.

[Splab]

You might want to go back to your books on that one. Yes they are all connected and one major issue is handling the amount of data going on the bus. Switching fiber is VERY expensive and therefore not something you are likely to find in a USB hub anytime soon.

Re:I'm more concerned with latency.

[eh2o]

Don't confuse latency (delay) with thru-put (avg bytes per second). Anyways, most USB-serial devices are CDC 1.1 or basically equivalent, and only full-speed at best (I have yet to see a high-speed USB-serial device). This means bulk transactions of 64 bytes each, so the host/driver/device has to deal with a pretty high interrupt rate to get the theoretical maximum. Long or poorly-shielded cables can also have a detrimental effect as bulk transfers are checksummed and assured delivery, so line noise causes the stream to stall and retry (unlike RS232). Bulk transactions also have the lowest priority on the bus so it should be expected that performance drops rapidly when other devices are present.

Re:I'm more concerned with latency.

[eh2o]

In order to achieve that low overhead firewire controllers require a significant amount of dedicated silicon which translates to a higher manufacturing cost. However with the proliferation of additional general-purpose and specialized cores on the modern CPU I expect the overhead problem will eventually be moot.

Re:I'm more concerned with latency.

[eh2o]

USB isn't a decentralized network like ethernet. There are no collisions because the host controller manages every transaction. This feature, by the way, makes possible a number of interesting things such as isochronous endpoints that have dedicated bandwidth and bounded latency.

Re:I'm more concerned with latency.

[raddan]

I've found wrt Firewire vs. USB, that USB tends to be "flakier" than Firewire. When doing long disk transfers via USB, I will occasionally lose the connection between the device and the host. Power interruptions (like AC ripple) seem to have some role in this, and putting everything in the signal chain on a power line conditioner has mitigated the problem somewhat. But I don't see the same behavior at all with Firewire, even with the same external disk (where the disks have both USB and Firewire controllers). I have no technical explanation for this, but my own anecdotal experience has led me to prefer Firewire for external disks. It might just be that there are more crappy USB chipsets out there, but it would be nice if USB3 addressed this [admittedly possibly imaginary] problem.

Re:I'm more concerned with latency.

[Hadlock]

Shit, in 2002 (120gb drives were just reaching the $150 range), you were lucky if you could buy a drive that would do 40mb/s (real world). Drive makers were claiming 44-48mb/s, but real world was 36-39. I'm not sure what it is today, but I can't imagine it's improved much since then. Even today, it's tough to max out a FW400 line connected to a single drive... it's just not the weakest link in the system... still.

Re:I'm more concerned with latency.

[|Cozmo|]

Yes it is true and will always be true for usb 2.0. Bulk transfers which is what a storage devices uses, cannot actually max out the bus due to the small packet size. Webcams on the other hand (high speed ISOCH transfers) can max out the bus easily.

It may be possible to create a storage device that multiplexes several bulk endpoints but I'm not sure how much faster you could get it.

Re:I'm more concerned with latency.

[durin]

Hmmm. This sounds familiar. Anyone remember the Fast Ethernet vs. 100VG fight?

Re:I'm more concerned with latency.

[FrankieBaby1986]

If transferring DV, and I want to transcode or do anything else (perhaps not related) then overhead dows matter. Why is it people feel they can use increasing processor speeds to avoid optimizing code and making things more efficient? Couldn't our computers be much faster than they are today, with speed optimized software) and yes, I realize that portability in code usually comes at a performance price.

Re:I'm more concerned with latency.

[Anonymous Coward]

in a pinch you could daisy chain PCs with it for an impromptu 400mb/s lan.

That works in Mac OS X, *nix, FreeBSD, and XP but Vista does not support it.

Re:I'm more concerned with latency.

[Gr8Apes]

And I'd rather not have a large disk to disk transfer interfere with my recording an HD channel.

Re:I'm more concerned with latency.

[Televiper2000]

The peer-2-peer is the real winner in Firewire along with the CAT-5 implementation (all the nice isolation you get). We built a synchronous Firewire device network with only a serial RS232 connection to the host. USB will always have it's place for being a cheap, simple, and robust next step for serial buses. The fact that it's port powered is another huge bonus. SATA is optimized as a close quarters serial bus. It can be bused down a cable but there's no point when USB already does the job perfectly fine.

Re:I'm more concerned with latency.

[LackThereof]

To be able to put a practical high performance RAID enclosure on USB? AWESOME.

Re:I'm more concerned with latency.

[Trogre]

Only problem was Apple screwed up by trademarking FireWire, so we get checkboxes such as iLink or IEEE-1394. No wonder nobody uses it today.

Re:I'm more concerned with latency.

[Emetophobe]

From the article you linked:

"As far as we know, effective rate reaches at 40MBps or 320Mbps for bulk transfer on a USB 2.0 hard drive with no one else is sharing the bus."

The part I emphasised is probably the most important part of that sentence. I've run read tests on both my external hard drives in the past and they maxed out at around 30MB/sec. I have around 4 other USB devices plugged in at all times, so that might explain while my externals are only getting closer to 30MB/sec over 40MB/sec. I always assumed that each USB device could get 480Mbps PER device, not 320Mbps for all devices... That's pretty crappy now that I think about it. I have 2 externals, a usb headset, a usb mouse, a usb gamepad, an ipod that uses usb, etc.. All these devices have to share that paltry amount of bandwidth? Now I understand why people prefer eSATA....

Re:I'm more concerned with latency.

[jhol13]

Instead of eSATA why not ExpressCard?

Re:I'm more concerned with latency.

[greed]

Don't forget, Firewire is full duplex and USB is half duplex. That is, the 4 pairs in "video camera" and "PC laptop" Firewire ports are TX+/TX- and RX+/RX-. The 6 pin connector adds 12-48VDC and ground.

USB has D+/D-, +5VDC and GND. D+/D- have to be reversed between transmit and receive, which means the devices on the bus have a "who talks next" protocol, just like good ol' unswitched Ethernet.

And we know from Communications in 3rd year Electrical that classic Ethernet saturates at 30% of the specified data rate, which is why switched Ethernet is such an important enhancement. (Not saying USB saturates at 30% of bus speed, but it will have a similar, less-than-100% figure.)

Really, the overhead of having to stop a stream to send a command packet is what kills USB.

That being said, getting the speed up by a factor of 10 means that such overhead does become smaller; though some issues I suspect will remain measured in wallclock time, not in functions of bus speed, so (say) read-to-write delay time may become a larger number of bits at higher speeds.

Re:I'm more concerned with latency.

[peter]

I can't find the article I saw, but the plan is to design a new driver interface along with USB 3.0. It seems they realize how much it currently sucks. Less CPU overhead = less power, too.

FP: And sometime in 2015...

[Tastecicles]

...a storage device that'll run at bus speed. What use is 4.8GBit if the attached drive bursts at 150MBit? Or is the USB RAID stack waiting in the wings?

Re:FP: And sometime in 2015...

[somersault]

Nothing wrong with being prepared. It's better than having an awesome device and knowing that USB is limiting your throughput. Can you imagine a beowulf cluster linked together by USB 3 cables, while running Linux, by our evil overlords in Soviet Russia?? Mmmmmm

Re:FP: And sometime in 2015...

[*weasel]

Because it's designed to serve -several- devices on that 4.8Gb bus and needs to not bottle-neck the much-faster solid state disks that will be commonly available in the next couple of years? To say nothing of handling the ever-increasing size and speed of digital still and vid cameras, scanners and print jobs of the current and near-future.

Re:FP: And sometime in 2015...

[jonatha]

Can you imagine a beowulf cluster linked together by USB 3 cables, while running Linux, by our evil overlords in Soviet Russia?

Administered by Natalie Portman wearing nothing but hot grits....

Re:FP: And sometime in 2015...

[somersault]

I am Natalie Portman, you insensitive clod!

Re:FP: And sometime in 2015...

[Braino420]

...a storage device that'll run at bus speed. What use is 4.8GBit if the attached drive bursts at 150MBit? Or is the USB RAID stack waiting in the wings?

2015? I think not. That thing must be unbearably slow w/ USB 2...

Re:FP: And sometime in 2015...

[SeaFox]

...a storage device that'll run at bus speed. What use is 4.8GBit if the attached drive bursts at 150MBit?

You would still get better performance from the drive since USB 2 only actually runs at 40MBit. The 480MBit is a burst rate. So USB 3 would (we assume) provide better sustained throughput maybe (if we follow the 10x performance claims) it would reach 400MBit, putting it level with Firewire 400.

Size?

[toppavak]

I wonder how adding fiber optic links will affect size and power requirements of USB3 devices. Granted, small LED's use minuscule amounts of energy, but wouldn't having to squeeze in power supplies and photodiodes at each end of the cable make it more difficult to squeeze it all into the micro-USB-sized interfaces used on most phones and mp3 players?

Re:Size?

[Khyber]

Well, for one, the photo-emitters would be in the devices themselves, in the port, not the cable. Same with power supplies. The day a CABLE needs a power supply is the day mankind has royally fucked up with technology.

Re:Size?

[mrchaotica]

They're not getting rid of the copper, you know -- they can still transmit the power over that.

Bottleneck?

[mrjb]

Currently I'm getting transfer rates of about 16 megabyte per second on hard drives connected via USB. That's roughly 160 megabit per second, whereas USB 2.0 can transfer up to 480 megabit per second. While I'm all for faster and better, the bottleneck seems to be elsewhere in this case.

Re:Bottleneck?

[hcdejong]

In the controller, likely. I'm getting 30% higher transfer rates with FW400 than with USB 2.0 on the same external disk.

Which doesn't give me high hopes for USB3. High-speed links are all good and well, but if they keep including cheap-ass controllers, what's the point?

Re:Bottleneck?

[MtViewGuy]

16 MB/sec. is not too bad if your external hard drive is used primarily as a backup drive.

Which does remind me: how come I haven't seen tape backup drives for sale that run through the USB 2.0 port?

Re:Bottleneck?

[Jah-Wren Ryel]

In the controller, likely. I'm getting 30% higher transfer rates with FW400 than with USB 2.0 on the same external disk. It's as least as likely the problem is in the protocol. USB is synchronous - data every packet received must be acknowledged in a return packet before the next data packet can be transmitted. That back and forth for each data packet means a lot of wasted time where the channel is essentially idle. Sometimes using a shorter cable can make a noticeable improvement.

Firewire has both synchronous and asynchronous modes. In async mode, a bunch of packets can be transmitted before any acknowledgment back is required. That's bad if the cables is flakey, since it will result in a lot of retransmits, but bad firewire cables are the exception, not the rule. So async is almost always way more efficient than synch. I'm pretty certain that you are using the async mode for talking to your disk.

Re:Bottleneck?

[iphayd]

This is common, and designed into the spec.

USB2's 480 Mbps was designed for marketing consumers, and as such is _burst_ speed. This way Intel could say "see, this number is higher than

Firewire's 400 Mbps was designed for video professionals, and as such is _sustained_ speed necessary to run video.

It's not a matter of chipset, it is a matter of protocol. Friends don't let friends get USB2 hard drives.

Re:Bottleneck?

[Gr8Apes]

You're not alone in that observation. It's why I've gone FW on all my external hardrives. The bus supports daisy chaining and no where near the anemic performance of USB 2.0.

Re:Bottleneck?

[Jeff DeMaagd]

USB is notoriously inefficient. From what I understand, they made some design decisions that prevent the protocol from operating quickly. This is why people find that Firewire devices end up being noticeably faster even though it's bitrate is 16% slower.

Re:Bottleneck?

[Immerial]

The other big problem with USB is that it requires the computer to act as the controller. Firewire allows for direct transfers between devices without the computer having to act as a relay.

Re:Bottleneck?

[Garabito]

As far as I know, the term 'Synchronous' refers to the use the same clock signal at both ends of the communication link to synchronize the data transfer; in that sense, USB wouldn't be synchronous.

I think the performance edge that Firewire 400 has over USB 2.0 has to do more with USB having a host/peripheral scheme, where the low level protocol operations rely on the host processor; while FW is more peer-to-peer oriented, so those I/O operations are carried by the device controller and the host controller; as a result, data transfer rates are not severely affected by the CPU load or speed in the latter case. Of course, this scheme makes Firewire chipsets more expensive.

Re:Bottleneck?

[Anonymous Coward]

How does incorrect shit keep getting modded up? USB is NOT just synchronous. USB has 3 modes: async, sync, and isochronous.

Re:Bottleneck?

[Jah-Wren Ryel]

As far as I know, the term 'Synchronous' refers to the use the same clock signal at both ends of the communication link to synchronize the data transfer; in that sense, USB wouldn't be synchronous. You are right, it was early and I didn't feel like digging deeply for the correct terms. However, my description of the behaviour of the protocols is correct to the best of my knowledge.

Yeah, but..

[theorem4]

Will this make the new cables more expensive?

Re:Yeah, but..

[LiNKz]

Also, if you integrate an element of fiber optics into a cable that routinely wrapped up, stepped on, or just basically abused, wouldn't it fail far easier than a standard cable?

Re:Yeah, but..

[jsiren]

You say that as if it were a bad thing. More failures - more sales...

This message was brought to you by the East Asian Cable Company.

Re:Yeah, but..

[walt-sjc]

I'm sure you be able to buy a Belkin or other generic for $20 or so, but Bad Buy will only carry the "Monster" models for $526 and up.

Re:Yeah, but..

[ggeens]

When I was working on my Master's thesis, I had to splice optical fiber a few times. Believe me, it's not easy.

Glass fiber is very flexible. You can bend it in any way you want, it won't break. You can cut it, but that takes considerable force. If you break the fiber, you'll break the copper wires as well.

Personally, I think the weakest point in such a cable will be the connectors. Getting the light from one fiber to another requires careful alignment. Any deviation might causes loss of signal. Getting dirt into the connector is probably fatal.

Re:Yeah, but..

[j33pn]

Exactly! Aren't fiber optic cables REALLY delicate? I don't think they can handle the abuse of a ubiquitous USB cord you throw into your laptop bag haphazardly.

Re:Yeah, but..

[Anonymous Coward]

"Exactly! Aren't fiber optic cables REALLY delicate?"

No, they aren't.

Re:Yeah, but..

[clickclickdrone]

>Getting dirt into the connector is probably fatal.
That sounds harsh. Do they actually send someone round to do a hit or do you get electrocuted or somesuch?

Re:Yeah, but..

[siwelwerd]

Hopefully they'll be less tasty than USB2 cables. My cats have destroyed 3 of them...

Re:Yeah, but..

[Anonymous Coward]

I don't think they can handle the abuse of a ubiquitous USB cord

"...that word. I do not think it means what you think it means."

Re:Yeah, but..

[David_W]

Getting dirt into the connector is probably fatal. That sounds harsh. Do they actually send someone round to do a hit or do you get electrocuted or somesuch?

No no no... remember, most optical connections use a laser. When you try to plug in the dirty connector, the laser reflects off the dirt toward you, instantly vaporizing you.

Re:Yeah, but..

[Fifty Points]

aren't we at war with east asia...

Re:Yeah, but..

[krakelohm]

Thats only on the USB 3.0 shark tank.

Re:Yeah, but..

[eh2o]

Nooooo!!! Sharks with frickin' USB3 ports on their heads!!

Re:Yeah, but..

[Big Boss]

No, we're at war with eurasia, and always have been.

RTFA

[Anonymous Coward]

From the article:

In addition, USB 3.0 will offer greater energy efficiency, Gelsinger said. It will be backward compatible, so current USB 2.0 devices will be able to plug into USB 3.0 ports.

RTFA instead of going for "First Nonsense Poster"

[Anonymous Coward]

Quote: "Gelsinger said. It will be backward compatible, so current USB 2.0 devices will be able to plug into USB 3.0 ports."

USB hard drives at 4.8Gbps?

[Gothmolly]

Um, what physical drive can push remotely close to that?

oh goodie

[maniac/dev/null]

Oh good. Now I get to plunk down $20+ per cable for the latest USB standard. I really like that copper USB2 cables are just about down to free from some online retailers. Looks like that will not be the case with USB3.

Re:oh goodie

[Waffle Iron]

You could get them from a big box store, where *all* cables come entombed in 3 layers of hermetically sealed PVC clamshell and cost $59. That way you wouldn't have to pay more for USB3 than USB2. The bonus is that the fiber link would be oxygen-free!

Re:oh goodie

[hcdejong]

I've bought Toslink cables for E 1.50 (one metre) to E 7 (10 metres), hardly more expensive than USB cables.

Re:oh goodie

[Just Some Guy]

Oh good. Now I get to plunk down $20+ per cable for the latest USB standard.

And you'll still be able to spend $79.99 for a 2m Monster USB 3 Premium AV Cable With Gold Plated Latinum Connectors. This is of course what the twit at Best Buy will insist on selling to your dad because it "makes the digital audio sound warmer".

Re:oh goodie

[izomiac]

Well, a quick check of Circuit City's website reveals that Joe Six-pack probably pays about $30 for a USB cable right now. The new cables are probably going to be ludicrously expensive. Or perhaps the price will stay at $30 since I'd think people would catch on to the scam when a USB cable costs more than their printer...

Re:oh goodie

[toddestan]

I wonder if that racket is about to run out on these retailers. Most people who are buying a printer nowadays probably already has a USB A to B cable from an older printer, and if they are a little bit smart they probably realize that they can use the same cable again denying Best Buy of their $29.50 profit on the cable sale. I bet that we'll see USB 3.0 printers soon, despite printers having no use for tha kind of speed. Don't expect a cable in the box, of course.

Re:oh goodie

[maxume]

You don't already have too many usb cables!?

The irony...

[packetmon]

In an interview after the speech, Gelsinger said there's typically a one- to two-year lag between the release of the specification and the availability of the technology,

In today's news, vendors worldwide urged one another to move quickly and get IPv6 deployed by the year 2025. When asked about the one or two year lag between the release of specs and the availability of the technology, vendors quickly pointed out the timeframe it took to implement Packet Over Bongo and IPv6 for Refrigerators. "It's been a long time in the making (IPv6) but we've finally succeeded in getting console connectivity to the fridge. We can now via a command prompt: finger lettuce" stated the happy refrigerator engineer. We never even knew of the existence of IPv4 for refrigerators. Engineers estimate another 20-80 year wait for IPv6.

Re:The irony...

[dorianh49]

Isn't Ubuntu making an OS, named Dumbuntu, especially for refrigerators? I thought the upcoming beta has been code-named Crispy Cucumber. I think I'll wait for E.Coli-laced Eggplant, though. Hopefully, it will be supported past the expiration date of my milk.

Plug Shape

[Crock23A]

Could be slightly off topic but I had to sound off on this one...

While I appreciate USB's capability for backwards computability, I would much rather have a plug shaped in such a way that I didn't have to flip it over every time I try to plug it in. I don't know about you guys but this is one of the most annoying aspects of using my computer, and I run Windows!

This would also be a great time to make a universal "other side" of the cable, rather than having a different plug for every single USB device. I have a mini plug for my camera, a big square one for my printer, a 2.5 mm jack to charge my MP3 player, etc. All these cables make a mess. If all my devices could share one cable, I'd be much happier.

Re:Plug Shape

[Aladrin]

While I'm mildly annoyed by having to flip them over, I quickly learned to simply look at the plug before I stick it in. Having it be flippable would mean duplicating wires and/or contacts and would make the cables more susceptible to damage and more expensive.

As for the ends... Blame the device manufacturers. There were originally 2 ends, the fat one and the flat one, and there was 1 of each on every cable. All the others with 2.5mm and other proprietary ends that work on nothing else are solely the fault of the OEMs. Nobody asked them to do it, it never made sense to do it, and it's just a huge pain in the ass for everyone. Even the 'mini' one that's often used on small devices now wasn't originally in the standard, it was added later since a device that's only 1/4" tall can't fit a 1/2" plug into it. (I don't know the actual height, but you get my drift.)

Re:Plug Shape

[Rob T Firefly]

All the others with 2.5mm and other proprietary ends that work on nothing else are solely the fault of the OEMs. Nobody asked them to do it, it never made sense to do it, and it's just a huge pain in the ass for everyone. It makes a whole lot of sense to an OEM who wants to be the only one who can sell their customers a cable despite an open standard. It's the same reason for every standard out there which suppliers have taken it upon themselves to add their own "enhancements" to.

Re:Plug Shape

[JanneM]

Could be slightly off topic but I had to sound off on this one...

While I appreciate USB's capability for backwards computability, I would much rather have a plug shaped in such a way that I didn't have to flip it over every time I try to plug it in. I don't know about you guys but this is one of the most annoying aspects of using my computer, and I run Windows! You know, you should only need to flip it over half the time. If you really need to do it more often than that, either your subconscious is an a*hole that likes messing with your self; or the world really is out to get _you_, specifically.

Sucks either way.

Re:Plug Shape

[Tribbin]

I like pluggin my USB-mice and USB-keyboard better than my PS/2 onces and I never broke an USB connector or anything so it's not that bad.

if !(plug()) {
    turn();
    plug();
}

If that is your biggest annoyance then I should give Windows a try again.

Re:Plug Shape

[Maximum Prophet]

Also, did you know that the USB connector is exactly as wide as an RJ45 network connector. No, neither did I, until my mother plugged her mouse into the ethernet port. (:-)

Seriously, the connector should have double the number of pins as it needs, and they should be symmetrical. That would also increase reliability, because if the cable didn't work one way because of a bad pin, just flip it over until you can buy a new cable.

Re:Plug Shape

[Bios_Hakr]

Didn't the people who "own" CDDA refuse to let DRM-protected audio discs be called "CDs"?

Why don't the USB people refuse to let the mini plugs be called USB?

I understand that it's impractical to have a full-sized USB plug on a tiny digital camera. But Sony should not be allowed to create a funky interface and then put "USB-Compatible" on the box.

Re:Plug Shape

[Eccles]

Having it be flippable would mean duplicating wires and/or contacts and would make the cables more susceptible to damage and more expensive.

I don't think the idea is to have the cable flippable, but instead to have some indication in the shape of which way around it goes. Firewire connects have a rounded end and a squared off one, for example.

Re:Plug Shape

[Anonymous Coward]

Every USB cable I've used as a USB logo on the top.

Re:Plug Shape

[Skapare]

Having it be flippable would mean duplicating wires and/or contacts and would make the cables more susceptible to damage and more expensive.

I disagree.

It is possible to make a connector that when rotated 180 degrees still fits the mating connector and still lins up contacts. It would be a "180 degree rotation symmetric" connector. Then you simply assign the contacts such that when plugged in either way, the same pair always contacts each other. The difference would be a reversal of data polarity. And that is irrelevant as most technologies work on edge-trigger or can be made to do so. Power would either require a rectifier bridge at the receiving end to handle polarity flips, or the power would be done with both contacts being the same polarity (added current capacity) and the shield be the power return.

Re:Plug Shape

[Just Some Guy]

That would also increase reliability, because if the cable didn't work one way because of a bad pin, just flip it over until you can buy a new cable.

Manufacturers: "it will be more expensive and we'll sell fewer of them? Ixnay!"

Re:Plug Shape

[Aladrin]

You're right. I forgot the way more complicated solution. Silly me!

Re:Plug Shape

[Skapare]

Maybe this would work for you.

I already have a plug design that, as one of its features, can be rotated 180 degrees when plugged in, and it still mates correctly for both data (separate pairs for each direction) and power (power can be supplied in either direction). The other feature is that not only is each end of the cable the same (at least for the same size class ... still need to allow for very tiny connectors for very tiny devices), the connector is asexual, meaning that it can mate with one of itself with just a 90 degree turn either way, with the data and power connections still made correctly (power polarity can be reversed, but a bridge rectifier can deal with that at the power usage end). Thus the connectors on the devices would actually be the same as the connectors on the cables. The only time you'd need a cable with different connectors on each end is when going between a large connector and a small connector (I think only 2 sizes would be needed, so this should be a minimal issue).

The connector would basically be square with locking clips on 2 sides, and holes on the other 2 sides to receive the locking clips from the incoming connector. Inside the connector square would be a pair of diagonal substrates arranged with the 90 degree angle corners pointing inward at each other. This would leave a pair of spaces for the mating connector to fit in. The data contacts would be on the sides close to the center. The power contacts would be on the same sides but further outward from the center. Data transmit would contact data receive on the other connector, with the polarity reversed if the connector is inserted at 180 degree rotation. But it would still be transmit connecting to receive (a total of 4 data wires). Power would similarly have 4 wires, 1 pair for power in one direction, and the other pair for power in the other direction. Devices not needing power just won't connect to the supplied power.

The only issue that you might have left is the issue of having normal size connectors (about the size of a device end of a standard size USB, or about the size of the 8P8C connector typically found on CAT5 cable), and tiny miniature connectors for devices too small to take a normal size connector. Cables could be made to have one size on one end and the other size on the other end, in which case you'd have to be sure you string up the cable with the correct end at the correct device.

By having the locking clips protrude from device mounted connectors by a slight amoung, maybe 0.5 to 1.0 mm, it would be possible to feel the connector hole to determine the alignment, whether the clips are horizontal or vertical relative to orientation of the device. Then just insert the cable at +90 or -90 degrees (either way will work).

Re:Plug Shape

[Paradise Pete]

While I'm mildly annoyed by having to flip them over, I quickly learned to simply look at the plug before I stick it in.

Or you could do what I do, put a dot on the side that goes up.

Re:Plug Shape

[iabervon]

Actually, the mini-USB connector is standard, and they made it to replace all of the proprietary ones. The proprietary ones were used because the regular square connector was too big to use on portable devices. People weren't going to be happy with a digital camera where the space in the device devoted to optics was smaller than the space devoted to the USB jack. On new devices these days, you pretty much only see the little flat connector on things you're expected to unplug and the big square one on things you aren't expected to unplug, and the important difference is how tightly they hold.

Re:Plug Shape

[Aladrin]

"wasn't originally in the standard, it was added later"

Yeah, I covered that.

Good point on the 'unplug' thing. It's definitely easier to plug and unplug the smaller one, especially since seeing the corners on the big one takes me a couple tries if it's turned funny.

Re:Plug Shape

[eh2o]

In spite of the standard it seems that the digital camera manufacturers still use their proprietary mini-plugs, which IMHO is rather annoying.

Re:Plug Shape

[MarsDefenseMinister]

Engineers: you can interleave the wires from one side with wires from the other side. Signals go on one side, and ground goes on the other. You can send data at higher rates down the now ground wire spaced wires, and justify your extra cost for the additional wires.

Sometimes it takes an engineer to think like a businessman.

Re:Plug Shape

[Skapare]

It's only more complicate ONCE for the engineer designing it (and only slightly more so). After that it's transparent to users.

Re:Plug Shape

[Grishnakh]

No, it's also much more expensive, because you're now talking about adding extra components, which will add lots of cost when you're making quantities in the millions. Just some extra diodes for your polarity-swappable power will add a few cents per board, which adds up to tens of thousands of dollars in million+ quantities. It's cheaper to just make the cable unidirectional and forget about the extra complexity, which almost no one will care about anyway.

Re:Plug Shape

[Grishnakh]

Why would they want to use standard plugs, when they can use proprietary ones instead, and charge you lots of money for replacement cables?

Frankly, I'm surprised any device manufacturer uses standard USB connectors at all.

Re:Plug Shape

[Anonymous Coward]

What bothers me most is that, in this day and age, these things still come in male and female connectors. This leads to a plethora of different cables, extensions, and connectors, and also a huge headache (made even bigger by your observation). What ever happened to hermaphroditic connectors?

Re:Plug Shape

[wfberg]

Let's say the pins are numbered as such:

1 2 3 4
5 6 7 8

let's say pin 1 is connected to pin 5, and 4 and 8 are never connected. Or maybe 8 isn't even connected to anything. The other end can simply test if there's a connection between 1 and 5; if not, the plug was flipped.
Worst case you waste 1.5 pins.

Same concept could apply to a circular plug that has pins arranged along its circumference; no way to plug it in wrong.

Re:Plug Shape

[iabervon]

Ah, yes. I spaced out in the middle of that paragraph and thought you were counting that in things you blamed on OEMs (the problem with reading slashdot while drinking your morning coffee...). The real issue with plugging and unplugging is how much strain the connector puts on the solder joint and the circuit board and the circuit board's mounting when you use it, and therefore how many cycles it can go through before something breaks off, especially when the device doesn't have room for a lot of physical support. How often the user fails to use it on the first try is somewhat less of a concern, although it's best if people always succeed on the second try, and don't keep pushing on the thing.

Re:Plug Shape

[Anonymous Coward]

Thanks for bringing this up! I first read about the pros of using such a cable in Jef Raskin's book The Humane Interface . He makes several interesting points regarding the ease of use and lowered cost (there is no need to design, manufacture, and stock all sorts of gender-bending adapters and extension cables) of gender-neutral connectors. I really wish those writing specs would put more thought into the physical interface to which users will be exposed.

Re:Plug Shape

[mollymoo]

Except USB uses differential signalling, so interspersing it with ground wires wouldn't be that much of a win.

Re:Plug Shape

[MarsDefenseMinister]

And that's why I'm not an engineer.

Re:Plug Shape

[Jeremy Erwin]

They are both part of the standard though mini-B is being deprecated in favor of micro-A.

Protected CDs can't use the CDDA logo because some standalone players (including, ironically, my late model Sony ES) can't play them.

Re:Plug Shape

[cthulhu11]

Congrats! You've just reinvented the IBM Cabling System.

Re:Plug Shape

[aliquis]

I wonder what the reason was to make a fat / flat one of almost the same size but no mini-one then.

Will this allow longer cables?

[Anonymous Coward]

Can I plug a USB hub into a motherboard on the other side of my house now without having to worry about signal loss or latency issues?

Eat into SATA?

[Random BedHead Ed]

Interesting. That should put the last nail into Firewire's coffin (and FW 800's). I wonder if we'll see USB 3.0 eat into SATA's market with internal USB3 drives on desktop and laptop machines. That could make desktops cheaper - ditching the IDE/SATA controller means one less component.

Re:Eat into SATA?

[hcdejong]

Let's hope so, it'd be good to see sanity return to the 'peripheral connections' market. FW400/FW800/USB2/eSATA is just too much.

Re:Eat into SATA?

[Gr8Apes]

Firewire blows away USB 2.0's performance for HDs in any meaningful benchmark you'd care to look at. eSata will probably replace it, whether that's a good thing or not. I don't believe you can daisy chain or use simple hubs with eSata, which means 1 cable/channel per logical device or expensive external controllers. It's certainly better than any USB concept due to the cluster that is USB. It's fine for secondary devices (keyboards, mice, graphic tablets, even phones) but not anything requiring high bandwidth.

USB is a lame horse that should be shot from the perspective of storage devices.

Re:Eat into SATA?

[stewbacca]

Last nail? They haven't even started fitting Firewire with a coffin yet. The only thing dying in this equation is USB2, and thankfully so. So it is faster than USB...joy, my mouse, printer and keyboard all now react the same speed, regardless of the interface speed boost. What's the point of USB2 again? If it doesn't boost the speed of my peripherals at least as well as my existing firewire400/800 ports, then why do I need it? It IS a pity that iPods stopped shipping with firewire and went to USB2, but that was a business choice to have maximum compatibility. Too bad my newer iPods can't transfer 2 songs per second anymore, like they used to when they were firewire400. But like many others have said, USB2 is "good enough" so it is really hard to get excited about USB3, which will be "gooder enough?"

great

[dropadrop]

I have a drive with USB2, FW400 and FW800 ports. In theory the USB2 should be a bit faster, but in reality it will transfer at about 25MB/s vs. 37MB/s for FW400. Strangely USB2 also consumes more CPU then Firewire. I guess now the time is ripe for USB3. With multicore CPU's we can dedicate one core just for use by USB.

Re:great

[Rocketship Underpant]

"Strangely USB2 also consumes more CPU then Firewire."

That's not so strange because one of the many ways Firewire is superior to USB is that each device has a hardware controller that negotiates data transfer over the bus independently of the CPU. USB, being a cheap-ass solution, relies on the CPU to do all that work, and is far more limited in a host of technical ways.

Re:great

[mr_jrt]

Though when you look at the originator of the standard, intel, it becomes quite apparent why it is CPU-bound. Want faster speeds? Buy a new Pentium! How I wish Firewire was the standard for all high-bandwidth connections. We might never have needed SATA nor USB 2.0 at all.

In such an event...

[Junta]

I'll wager a broken fiber in a cable would manifest itself as 'USB2 only' connection.

More half assed implementations.

[deniable]

Fiber and copper. Let's see how vendors screw this one up. USB 1, 1.1, 1.?, ?.?, 2, 2.75, 3...

Re:More half assed implementations.

[Anonymous Coward]

The implementation isn't as bad as the naming convention.

Do other people have difficulty remembering which is faster, full speed or high speed?

Since they are adding optics maybe it is time to scrap the old nomenclature and adopt the following:
sub-light speed, light speed, ridiculous speed, and ludicrous speed

Faster, possibly

[Tim C]

Although it'll depend on where the bottleneck actually is, of course.

But easier? How would it make using external HDDs easier?

Re:Faster, possibly

[Anonymous Coward]

If you tell people that don't know any better that something is easier, they will buy it.

Doubtful

[Anonymous Coward]

SATA is still better from a technical standpoint. 4 standard wires, no fiber optic needed. Plus SATA 6 Gbps will be out sometime here. SATA does have those crappy connectors though.

USB connectors are more robust. However, that robustness will be offset by their poor decision to use fiber optic cables. Fiber cables are a pain in the ass because they break easily. Even the slightest kink in a fiber optic cable will ruin it. Accidentally rolling over a fiber cable with your chair will usually kill it. The fiber might let you go long distances though, maybe.

Re:Doubtful

[Random BedHead Ed]

Yes, I was thinking about the fragility of the cables, but for internal components like a desktop hard disk it wouldn't matter all that much. Maybe in five years we'll be buying desktops with a terabyte of storage spread across USB 3 solid state hard disks in RAID 5. That would be fun.

I for one. . .

[Blinocac]

welcome our new fiber optically aided Over Lords.

Not in practice.

[pla]

This should make USB hard drives easier and faster to use.

Faster, yes. Easier?

Perhaps I count as an extremely unlucky outlier in my experiences with USB in general, but I have found it one of the buggiest PC interfaces ever. And I include VL graphics cards in that list.

XP and a modern machine finally seem capable of handling simple things like USB1 keyboards and mice properly. Printers, still asking for a reservation at the sanitarium. Cameras, not too bad, but they only need to work for five minutes at a time.

But HDDs... I've dealt with four different models on three different PCs running four different OSs (yes, four OSs on three PCs - I actually reinstalled three different Linux distros, using 2.2, 2.4, and 2.6 kernels, just to recover data I had trusted to an external USB HDD). And they all have exactly the same problem - They randomly drop offline just when you start hitting them the hardest.

Under XP, they appear and dissapear on a whim. Some days you can't even get them to stay connected long enough to format them, while others (rarely) you might have it work all day long.

Under Linux, I've had what you could technically call a better experience - As long as I limit it to UHCI (ie, slow old 12mbit USB1.1), it works great, rock solid. Toss in the EHCI driver to allow a decent transfer rate, and it zips along nicely - For about 45 seconds, after which it decides to offline itself until a reboot (even removing and reinserting the module doesn't let you bring it back up).



So... Forgive my skepticism, but the thought of a newer-faster-buggier-than-ever version of USB just doesn't get me all that excited. I think I'd use the phrase "fills me with dread".

Re:Not in practice.

[Boinger69]

I suffered the same issues with USB2.0, decent transfers for 30 seconds until it stops responding and the kernel resets the bus. Dropping back to USB 1.1 worked well, but in the end i just trashed the idea of USB external devices and went with firewire. So far everything works much faster than USB and takes alot less CPU for simple file copies and other general purpose use.

This goes back on an earlier poster's comment on the controller chips being cheap crap. 100% Truth there, every usb disk device I have seen has a cheap controller completely reliant on the USB host to baby sit and spoon feed it every operation to perform (I beleive this is by spec). Firewire on the other hand does not have this issue, the controllers are intelligent and are perfectly capable of controlling data transfer on their own.

Think PATA/SATA vs SCSI.

Re:Not in practice.

[domatic]

You've been unlucky. I routinely use external USB hard drives for backup with good speed and reliability. Perhaps you commonly deal with a flaky mobo chipset?

Re:Not in practice.

[cnaumann]

Same opinion here. Some printers work fine plugged directly into the PC but will not work when plugged into a hub. Same printer works, but only locally, printing over the network causes USB errors. Some devices are perfectly happy plugged into a hub, until you plug another similar device into the same hub, then after a while neither device works. (Not a voltage problem, I checked that.) I have had USB hardrrives pretent to work until I went to retrieve the data and found the entire drive corrupted. (And yes, I had deleted my local copies). I have seen notebooks that simply would not work with certain flash drives, and no one could tell me why.

I have built USB devices, and they work, sometimes, under certain conditions (mostly when they have a port all to themselves). Plug a bunch of them into a USB 2.0 hub and you can forget it, but a (hard to find) USB 1.1 hub works OK (albeit slow).

And as other posters have pointed out, the one-sides flat plug is not well though out, the "other end" is pointlessly far from universal, USB extension cords are "illegal" but still quite common...

It has been over 10 years now and USB is still buggy as hell. It makes me long for the old days of the old Centronics printer connector. Once you got past some basic.. finding a free interupt, the bi-directional settings, the correct "lp" numbers... those things always worked. At this point, I think USB is a do-over.

Re:Not in practice.

[spaceyhackerlady]

I find removable storage a pain on XP. I find XP a pain, generally. YMMV.

On my Mac and my Linux boxes external hard drives work just fine. On the Mac I have a choice of USB 2 and FireWire. Both work fine, though FireWire may have a performance edge. On my old Slackware laptop it's USB 1.1 or FireWire. Since my laptop only has a 4 pin FireWire connector I have to use an external power supply on external hard drives. This may help.

...laura

New cables (expensive?)

[amaiman]

This is going to mean you need special "USB 3" cables, which users will confuse with regular ones...

I assume that the cables will be much more expensive, as well, because of the fiber component. I can get a regular cable for about $3 now, does anyone know how much the new cables are likely to cost?

Re:New cables (expensive?)

[YrWrstNtmr]

does anyone know how much the new cables are likely to cost?

No.
FTA: "USB 3.0 products should likely arrive in 2009 or 2010."

Re:New cables (expensive?)

[Maximum Prophet]

If they design it right, using the wrong type of cable will only degrade performance, your device will still work at USB 2.0 speeds. It's also likely that your computer will negotiate with the device so that if the device says 3.0 but there's no signal on the optical link, the computer will tell you. Any system should be designed to fail gracefully.

Re:New cables (expensive?)

[magarity]

your device will still work at USB 2.0 speeds...the computer will tell you
 
Frankly, it is most irritating that XP on my old laptop with USB 1 complains that the device could be faster and offers to show me a list of ports so I can pick a faster one. WTF - I already know there is no USB 2 port in the thing. How many tech support calls are made when less technical users get this? It should suggest I go get an adapter but the list of ports is worse than useless; it implies that one of the other ports will do the job.

Sounds like...

[Tastecicles]

...you have a hardware problem. Check your power supply, you may be suffering what we in the trade call a +5SB undervolt.

Re:Sounds like...

[Maximum Prophet]

I'm having the GP's problem with an external Firewire drive. The power supply is where I'm going to check next, but I also suspect the connector to the PS is flaky.

Re:Sounds like...

[pla]

...you have a hardware problem. Check your power supply, you may be suffering what we in the trade call a +5SB undervolt.

The same problem with a racked PowerEdge server (RHEL), a top-of-the-line Latitude laptop (XP), and a hand-built greybox (XP, Slackware, Knoppix) using some of the highest quality parts available (ABit KN8-SLI, Seasonic power supply, OCZ memory)?

I would tend to agree with you for a one-off problem. But those three have literally nothing in common.

Like I said, perhaps I just have bad luck with USB. But I can only let it burn me so many times before I conclude the interface itself (or at least, most implementations thereof) sucks rather than merely my experiences with it.

Re:Sounds like...

[Tastecicles]

Could be, my experiences with usb problems lie with OTS boards - the USB bus draws power from the +5SB rail, which isn't good if it's a server environment. The more expensive (read: better quality) PSU boxen and boards keep the +5SB and the +5USB rails seperate. BTW, the laptop has only one +5 rail, and that's nothing to do with ACPI (which runs off a 3.6V Li-ion cell); it's purely and only for USB. The Abit KN8 is an off-the-shelf workstation board, chances are the SB and USB share the same rail. Not sure on the PowerEdge, some use desktop boards, some use 1U server boards.

Re:Sounds like...

[toddestan]

Perhaps a problem with the USB drive itself? I have a bunch of USB harddrives, which I put together using older drives fitted with generic no-name cases, and never had a problem with the USB connection on any of them. The newer cases have firewire, and that has given me trouble with one laptop, which is probably a fluke of the laptop given they work elsewhere.

Screw bandwidth...

[Balinares]

Give us a standard that actually delivers enough power that you don't need an additional power cord for just about every other device already... :/

Re:Screw bandwidth...

[wild_berry]

That's what the fiber is for: no E.M. interference. Then you can boost the power lines as needed for the device's optical circuits.

Re:Screw bandwidth...

[jimicus]

The problem with that is that a PSU for a powered 7-port USB hub already has to supply 3.5 amps. Much larger than that and with current technology you're getting into "big expensive PSU" territory.

Re:Screw bandwidth...

[cerberusss]

Give us a standard that actually delivers enough power that you don't need an additional power cord for just about every other device already... :/

According to Intel, the new USB 3 standard will use fiber optic cable for data as well as power. The data will be modulated on a high-powered laser light signal, enough to deliver the power to spin up a harddisk, or, alternatively, burn through one solid oaken office door as well as the sales guy who was about to open said door.

Re:Screw bandwidth...

[Skapare]

Or give us devices that don't need to consume so much power (and thus don't get as hot).

Re:Screw bandwidth...

[Skapare]

According to Intel, the new USB 3 standard will use fiber optic cable for data as well as power. The data will be modulated on a high-powered laser light signal, enough to deliver the power to spin up a harddisk, or, alternatively, burn through one solid oaken office door as well as the sales guy who was about to open said door.

Make that be 532nm wavelength light and we can use these in place of Wicked Lasers :-)

Re:Screw bandwidth...

[Jesus_666]

I'm positive that they will invent a 2.5" HDD that requires less than 0.5 amps to spin up soon. Maybe making the platters from unobtanium might help.

Re:Screw bandwidth...

[Skapare]

Microdrives already work. Initially they were made for Compact Flash (IDE compatible) slots. This is doable for USB drives.

The current crop of portable USB drives are still physically a bit large. They can be made smaller as drive capacities go up. Try a 1.8" hard drive.

Re:Screw bandwidth...

[Jesus_666]

Well, 2.5" happens to be what's currently popular - and, quite honestly, I'm not interested in smaller portable hard drives unless they happen to have the GB-per-Dollar rate as the bigger ones. Which, of course, is complete nonsense.

Microdrives and sub-2.5" HDDs are fine and dandy, but two dollars per gig are as unacceptable as the tiny amount of space you get (I haven't been able to find an 1.8" drive with more than 60 gigs; stuff like the Spinpoint N2 isn't readily available). Give me a 200 GB 1.8" drive or microdrove for 120 Eurobucks and I'm interested.

Since such small HDDs don't suit my needs at all I don't see why I should pay extra just to accomodate for an ill-conceived inter-device connection standard that never was revised in a place where it should have been. If instead I just pay twenty bucks more (still staying way below what I'd pay for the same amount of storage in 1.8") and buy my external hard drive with Firewire, all of the energy problems disappear and I get faster transfers, too. It seems that in my case the inadequate component is clearly the USB standard and not the hard drive.

Re:Screw bandwidth...

[Monkier]

we also won't have to go out and buy some non-standard power-pack when the original goes kaput.

Re:Screw bandwidth...

[Anonymous Coward]

Now you end up with a broken oaken door? Oxen boxen!

Re:Plug Shape (SCSI was worse)

[rubberbando]

It could be worse, we could all be forced to use SCSI devices instead. That had a ton of port styles and shapes as well as internal and external versions of each. Also, you had to daisy chain your devices, secure them with a screwdriver/your fingers, configure each with its own ID number and make sure the end of the chain was terminated properly. Then to top it off, cables, adapters, and terminators were insanely expensive. I'd swear that every time I bought a new SCSI device that it was like playing a puzzle game like Tetris with cables/ports instead of block in hopes that I wouldn't have to go back to the store and shell out $30 or more for an adapter or cable. Sheesh. I don't miss that at all... :P

USB 3.0?

[SlashDread]

Sure it isnt 1.3?

Om a serious note, why are E-SATA connects so impopular on new systems? These are a godsend for external drives, compared to any USB.

Re:USB 3.0?

[sayfawa]

This article sent me on a reading binge about all these different specs and ways to connect and yeah, it seems like eSATA would be the obvious choice. Maybe SATA is only for hard drives and not for things like flash sticks? I don't know.

Re:USB 3.0?

[pklong]

Because USB is a cooler buzzword than SATA ;)

Both Firewire and ESATA (with hot plugging drivers that actually work) completely outclass USB2

Re:USB 3.0?

[ElberethZone]

ESATA is not self-powered, you need another cable or external power. Appart from that, I don't know is SATA is usable for other devices that are not HD/SSD/CD/DVD.

Re:USB 3.0?

[Catbeller]

'cause USB was invented by Intel at the same time Firewire was put out by Apple (around 1999), and Intel became very good indeed at manipulating the motherboard market (using the old tactics like discount or supply refusal to manufacturers) to freeze out Firewire. eSATA I assume is the same story? USB will always dominate because Intel dominates. And Microsoft goes along with Intel; they draaaagggged their feet bringing Firewire to Windows. Ya seriously think that a piece of silicon controlling a firewire port costs more than a USB controller? These are parts that cost a couple of dollars. Firewire is Apple's baby, and even tho Apple charges nothing for licensing, political pressure became a marketing factor even to this day. Even tho the price difference to manufacture is nonexistent, the market droids have slotted Firewire as a high-profit niche product and aren't about to budge, 'cause they'd lose money cutting the price. eSATA, dunno about the politics there, but I'd guess it's much the same story.

Re:USB 3.0?

[19thNervousBreakdown]

Neither is USB when you're running a hard drive.

Re:USB 3.0?

[ElberethZone]

Yes it is... I have external enclosure for 2.5 inches HDD and it is self powered from USB. USB power is just not enough for 3.5 inches one but not for 2.5 or SSD.

Re:USB 3.0?

[19thNervousBreakdown]

No kidding, I stand corrected. Danke :)

Why not converge all of the bus standards?

[Theovon]

I wonder if eventually the speed and latency of USB will reach a point where SATA, for instance, becomes unnecessary. Or how about ethernet? We'll always need REALLY fast links like PCIe or dedicated ones like DVI. But when it comes to busses, perhaps it would be good to settle on one standard. I'm envisioning something along the lines of 100 gigabit ethernet, except all computers would have a whole bunch of dedicated ports, rather than just one network.

Re:Why not converge all of the bus standards?

[Corporate Troll]

We'll always need REALLY fast links like PCIe or dedicated ones like DVI. But when it comes to busses

You do realise that PCIe is a bus, right? (I think DVI is also a bus, but I'm not 100% sure if it matches the traditional definition of a bus)

Re:Why not converge all of the bus standards?

[Skapare]

A converged bus would be a compromise in a lot of ways. But I do believe it is feasible for a large portion of device connectivity, including just about everything now connected via USB, FW, or SATA. PCIe and DVI would pose some complications, but video up to a point could be handled over a single data pair (DVI has separate pairs for red, green, and blue). SATA shows us that we can do 3 GB data rates on metallic conductors. One problem is that all the connectors are bad designs. They are flimsey. They require exact orientation. And they have both male and female ends. My connector design would eliminate those issues.

My connector design would be a square connector (about the size of the device end of USB) with locking clips on 2 opposing sides, and holes to accept the locking clips on the remaining 2 sides. The design is to allow one connector to mate with one exactly like it rotated at 90 degrees. Inside the square would be two plastic parts shaped in diagonals with the 90 degree corners pointing at the center and at each other. The space between the diagonals would be where these parts from the incoming connector would fit. Think of the letter X where one connector has these plastic parts in the upper and lower gaps of the X and the incoming connector turned 90 degrees would fit with the parts on the left and right gaps. The contacts for data would be near the center on the sides of the diagonal parts. For data transmit, one contact would be on one of the parts, and the other contact would be on the other part at 180 degrees. So if you rotate the connector 180 degrees, you are only flipping the polarity of the wires making up the data transmit pair. The data receive pair would be on the other side of the parts, also near the center. Power (deliverable in both directions) would be on the same sides but further away from the center. The power polarity would have to be corrected with a bridge rectifier in the device taking power. Polarity on the data wires would be a non-issue since modern data encoding doesn't need this (e.g. edge-triggered, etc).

The middle of the diagonal parts would be reserved for future fiber usage with as many as 4 fibers between devices.

Using multi-level encodings, data speeds of 3 Gbits per second and higher are easily possible with metallic. While there would be separate transmit and receive pairs for simpler devices, the protocol could include a means to negotiate uni-directional bursting where at times both data pairs could be used at the same time in one direction to double the transfer rate in a specific direction. This could help handle uncompressed video to a monitor (while still allowing enough return bandwidth to let you plug the keyboard and mouse into the monitor, along with a touch screen sensor if you want.

The only real hindrance to converging all these technologies is people ... those who have some vested interest in each of the others.

Re:Eat into firewire not likley

[TRRosen]

more speed is great but USB will never replace FireWire for video as its not a matter of speed but architecture. FireWire is a true bus USB is not...FireWire can guarantee bandwith USB can not.

Yes, but

[Doug Neal]

Will the whole bus slow down to 1.5Mbit when you plug a mouse in? ;)

How exciting

[FileNotFound01]

This will come in handy when that 100,000 rpm USB drive I'm waiting for ships. It'll be faster and [even] easier to use, especially with the optional depleted uranium housing.

Better support as well as speed?

[BloodyIron]

So.... are they going to fix the bugs with USB 3.0?

Well, not really bugs, but things that need improvement.

IE: Not having to re-configure a USB device when plugging it into another port (on my motherboard, this problem does not exists on USB hubs)
IE: Windows XP support (lol, I couldn't help myself)
IE: Actual, intuitive networking, instead of a networking design which (to this day) few know how it works? If it says you can network computers with it, make it easy. Please, make some sort of (at least) cross over cable between two computers.

I'm sure there are other bugs, but I just dont know them :/

Re:Better support as well as speed?

[Chirs]

The whole thing about reconfiguring a USB device when changing ports is an OS thing, not a fault of USB itself. Linux handles changing ports just fine.

As for the crossover cable...have you ever bothered looking? For instance...

http://www.datapro.net/products/usb-2-0-host-to-host-cable.html

So we should be able to boot from a USB flash

[Rooked_One]

drive? It would only make sense that since its solid state it would be faster than our primitive hard drives with their moving parts... (yes, I know about SS Hard drives and don't have 2500 dollars to spare)

Re:So we should be able to boot from a USB flash

[Aeiri]

I've been booting from my USB flash drive for 3 years now...

Re:So we should be able to boot from a USB flash

[Rooked_One]

right, but will it be FAST as an internal drive? If its got the theoritical throughput... hmmm

speed over fiber vs speed of electrons over copper

[Anonymous Coward]

"10 times as fast by adding fiber-optic"

Someone explain how this is anything other than marketing.

As there is 10gigabit Ethernet over copper, why would 4.8 gigabytes require optical?

Why not just add a more durable plug to Ethernet, and just use it for everything? replace usb, replace DVI/HDMI ...

The speed is OS dependent

[Anonymous Coward]

with usb 2.0 hardware and drivers, Windows can get some pretty decent speeds with big hard drives, but Linux invariably falls back to 1.0 and the best rate I've seen with 2.6.any kernels is about 100kbs which translates into about 5 times dialup speed. This is somewhat below usb 2.0's theoretical capabilities, and the last backup I was able to make using linux, it took a week to backup 5 gigs to an external drive.

Re:The speed is OS dependent

[Anonymous Coward]

Bullshit. While I can't write to my external hard drive quite as fast in Linux as I can in Windows, that's because the drive is formatted in NTFS. It's still somewhere in the region of 10MB/s (probably more - I never measured it exactly), which is well above the maximum of USB 1. Copying 5GB of data onto it only takes a few minutes.

Now, either your PC or your drive may have some other weird problem, or you're using a Linux distribution so old it doesn't support USB 2. If your USB 2 hardware actually requires separate drivers in Windows, then chances are it's one of those weird non-standard ones. All the compliant UHCI USB controllers (the ones that work with no extra drivers in Windows) work in Linux too.

Re:The speed is OS dependent

[MarsDefenseMinister]

The reason you're getting the slow speeds under Linux is because you're not compiling your kernel properly. The OHCI and UHCI USB drivers are for the slow chips, and the slow speeds of the USB 2.0 chips. If you want the 480MBPS speed, then you need to compile the EHCI USB driver.

Re:The speed is OS dependent

[MBHkewl]

BULL!

I connect my external USB 2.0 HDD (VFAT formatted) and I achieve a max stable speed of 50 MB/s, sometimes up to 60 MB/s.
While on Winblows, I don't get speeds as fast, because if I copy the same files to the disk, it takes more time.

The end of HDMI?

[imrec]

How many more revisions before USB catches up with HDMI? (~10 Gbit/s?) Though not a big fan of USB, I DO like standardized components.

Anyone else here see a problem??

[zzsmirkzz]

You know I've been thinking, which I admit may be a bad thing this early in the morning. But, I've noticed more and more devices using USB as not only a data link, but as a power source. Some of them, only as a power source, or so it would appear. Now what's got me bothered, especially as they ramp the bandwidth of the data-transfer up, is the possibility of every device (Cell Phone, Printer, MP3 player, etc) spying on you and transmitting the data back through the power grid. I know its a super-conspiracy theory involving just about everyone except you, but, it still could happen, right?

Re:Anyone else here see a problem??

[aadvancedGIR]

Then I recommend you not to buy the USB-powered dildo.

Re:Anyone else here see a problem??

[mkiwi]

Then I recommend you not to buy the USB-powered dildo.

Stick with the one that plugs into the dock connector of an iPod. It's so much more portable and fun that way.

stop dropping video frames?

[DriveDog]

Will USB 3.0 be suitable for transferring video, etc? Currently I have to use Firewire to avoid dropped frames and such. I think it has little to do with USB 2's maximum rate.

Re:stop dropping video frames?

[stewbacca]

It's my understanding that USB is not a suitable video transfer technology because it doesn't send and receive data asynchronously like firewire does (or some terminology like that). USB is a one-way road whereas Firewire is two. Also, I've read a lot of technical mumbo-jumbo (that flies right over my head)about Firewire's cability to mark and catalogue digital film for proper frame alignment and "stuff"....ok, I'm done talking out of my a$$ now, I'll turn it back over to the experts.

Honest Question

[martyb]

I've got a question that has been nagging at me for quite a while and was hoping someone here could phrase an answer in terms a mere mortal could understand.

Why are there so many serial specifications?

We've got, off the top of my head, SCSI, USB, Ethernet, FireWire, and SATA to name a few. I do understand there are different protocols (all the way up from the physical to the application layers). Different applications of these technologies permit some optimizations that might not be applicable in other situations. But, at some point, the underlying technology is fast enough

Still, I can't help but think there should be some common denominator that ALL these communications standards can agree on, and through economies of scale, become universal standard(s). It just seems like people keep re-inventing the wheel with an eye toward THEIR favorite.

I thought we were getting close when they released gigabit Ethernet over UTP (unshielded twisted pair).

• can handle distances up to 100 meters
• fast data rate (1000 Mbps)
• supports lower data rates (100/10 Mbps)
• development is underway for 10Gbps, too.

So, for the sake of argument, why not have all of our serial devices just support gigabit Ethernet? Sure, you'd need a hub or switch in your PC to talk to all of the devices, but you already need something similar for the other protocols (USB hub, SCSI controller, etc.). It's a well-known technology with many implementations and is widely available. I'd willingly pay a few more bucks for each device if I could ditch all of these incompatible formats and just standardize on one SET of ports and cables for hooking things to (and within) my PC. And in those cases where a different connector is desired (e.g. for small form-factor devices like a digital camera), let me just get an adapter cable/plug that I can plug into my Ethernet port.

Is there any good, technical reason that is keeping us from having truly UNIVERSAL serial communications?

Re:Honest Question

[mlBrianR]

I think the main problem is repeated insertion/removal vs. semi-permanent installation.

USB connectors are designed to be inserted and removed over and over. They're held in by pressure against the connector, so they can be removed without having to push a tab or twist the connector to remove it.

UTP cables are designed to be plugged in, and then generally left alone. The UTP cable in my computer bag is in terrible shape.. the RJ45 connector is coming loose, the plastic retaining tab is broken off (so the cable often pops out of the jack on its own), etc.

I have USB devices which I've removed and inserted hundreds of times, and the connectors still work reliably.

Re:Honest Question

[rengav]

Is there any good, technical reason that is keeping us from having truly UNIVERSAL serial communications?

Other that physical size of connector, I can't really see one. There are many times you want a very small connector. Sure you could make a myriad of connectors for a common bus architecture, but then you're back to the gripe of too many f-ing cables. However, I can see economic and marketing reasons to have multiple standards all over the place.

Re:Honest Question

[rhadc]

It's not that simple. Why do we need so many models of cars? Needs are different between people and applications. Our applications will always develop into using 100% of the bandwidth provided.

The standards are developed at different times, and they have their own lifespans. RS232 is falling out of use, but it will be a loong time before it is gone. Firewire found new competition with USB 2.0, but if USB 3.0 is still asynchronous, it still won't compete with isosynchronous firewire's ability to guarantee a level of bandwidth on the wire. Addressing is another issue - some solutions allow for dynamic addressing, some for world-addressibility (IP by way of Ethernet), and some require manual addressing (SCSI, usually). Some protocols allow anything to be communicated (IP, Ethernet), some don't (SATA). The list of differences goes on and on. Over time, there may be consolidation of interfaces at the consumer level, but forget about the world finally arriving at one any time soon.

Re:Honest Question

[skives]

Ethernet is all signal, no power. You couldn't charge your iPod or iPhone with your sync cable, nor would your mouse or keyboard work without batteries.

It's a nice idea though... Ethernet along with the usual IP and TCP/UDP slapped on top are very mature standards that are easy to implement and troubleshoot. The recent debacle rolling out huge SAN networks at my company (for Sun servers) has me rethinking the viability of NAS just for those reasons.

Re:Honest Question

[smellsofbikes]

A bunch of other people have chimed in with parts of the answer already, and I thought I'd add some more.
Obviously each manufacturer wants you to use their standard and buy their hardware.
Different implementations came through at different times, and have different amounts of software/hardware overheads, and try to do different things.
RS232 has almost no necessary software overhead -- you do any and all the work with code you write. USB has *quite* a bit of software overhead to deal with device identification, and Ethernet has *enormous* amounts of overhead. In most small systems you have to buy an Ethernet software stack separately from the OS you're using.
USB tries to provide power. People are trying to glue power into Ethernet although it hasn't yet taken off.
People keep going off in odd wireless directions.
The fundamental problem, I think, is that there are several different connectivity needs and manufacturers are trying to get you to buy their solution to what they think are the most important needs. What you're asking for is something good for the industry in the long term, and that's not really in the direct interest of any particular company, so nobody's building anything for it.
The military embedded market seems to be moving towards gigabit or 10-gig fiber ethernet for all their interboard communications, but fiber has its own problems, and I'm not sure it's the right thing for a USB key you're carrying in your pocket all the time.
I believe that the SCSI module in linux handles firewire and USB, so from that standpoint it looks like it's a start towards universal communications, except for Ethernet. (Even though old SCSI is nothing like serial: it's the the ultimate expression of parallel communications, with some similarities to the old HP/GPIB parallel communication standard that's still used in for test communication but used to be a hard drive standard.) I have no idea what Windows does.

Re:Honest Question

[gknoy]

So use a USB-style connection for plug durability, and The One True Data Transfer Protocol (e.g., the GP suggests gigabit ethernet) to send the bits.

Re:Honest Question

[gillbates]

Is there any good, technical reason that is keeping us from having truly UNIVERSAL serial communications?

Yes.

Let me explain:

• USB, per standard (host side), must be able to source at least 500 mA. Even though there is a PoE (Power over Ethernet) standard, most choose not to implement it by default. Hence, ethernet can't power devices like flash drives and hard drives.
• USB uses four wires, ethernet twice as many. USB is a synchronous bus, meaning that there are (theoretically, at least) no collisions. A properly operating USB device will not stomp on someone else's data packet. Thus, for the given bitrate, a higher portion of the bandwidth is available to applications. Unlike ethernet, adding devices to the same physical connection will not degrade the overall bandwidth of the network. In practice, I've found 100 Mbit ethernet devices operating with a maximum throughput of about 35-40 Mbits/second because of collisions. And this was with *two* devices! To get a better throughput requires using routers (which minimize or eliminate collisions).
• RS232 is a pretty universal serial communication standard. However, it is also slow.
• There are tradeoffs between maximum cable length and the speed of the bus.
• There are tradeoffs between the number of signal lines and the cost of the device.
• There are tradeoffs between bitrate and the device cost.

So, the reason why we don't have a universal serial standard is because the different interfaces were designed with different goals in mind.

Re:Honest Question

[Yvanhoe]

Apple was thinking along these lines when they proposed firewire. I am not a Mac expert, but from what I have seen so far, firewire can be used for networking (does it use ethernet, I am not sure) video transfer, peripherals connection, external drive connections. From the start it was fast enough and designed for all these applications while we still struggle to get USB working everywhere at full speed...

Re:Honest Question

[karnal]

USB tries to provide power. People are trying to glue power into Ethernet although it hasn't yet taken off.

I take it you've not implemented a large-scale VoIP environment, or installed an enterprise with Wireless Access Points. Sometimes POE does have applications, although admittedly away from home or small business....

Re:Honest Question

[SleepyHappyDoc]

I was thinking a similar thought, but backwards. Why don't they implement whatever Ethernet requires, over this USB 3? Properly designed, this new USB could replace the Ethernet connector in home desktops.

Re:Honest Question

[smellsofbikes]

I work for a company that designs and supplies PoE driver chips, so when I say "it hasn't yet taken off" I'm talking about what people were forecasting for sales three years ago, what we're actually selling, and the problems we're seeing with consumer acceptance of our parts, particularly when a particular very large company who has a dominant position in the industry and has packed all the standards committees with its representatives is selling large quantities of equipment that claims to meet industry PoE standards but actually only works with other members of its own equipment.

Re:Honest Question

[karnal]

I'd have to take a wild guess that the company you're referring to starts with a C and ends in a disco (without the d.) I've seen that solution fully implemented, and I'm in the middle of implementing another company's standard - of which I'm told follows a better "set" of standards than Cisco. Of course, the switches and phone vendor are partnered to work together... and I honestly haven't dug into any of the official standards for POE to see if they can power anything else at all.

So I would have to agree with your statement regarding things not taking off. Especially if everyone decides to go gig - if your device supports gig copper, I'm not quite sure how it can use any POE standard, unless end devices can be smart enough to ignore an elevated level of power on some of the pairs - and it would have to be every device out there...

Re:Honest Question

[smellsofbikes]

You deserve a +3 insightful for that group of observations.
What sucks is that consumers are buying things that claim to be standardized, and they don't work together even though they all list compliance with the same standard. That's an awesome way to piss off customers and destroy any trust anyone might ever have in the technology.

Re:Honest Question

[adolf]

I'm all for the USB connector's reliability. Myself, I even go through the trouble of using a cheap USB Ethernet adapter whenever I'm in the field, so that a somewhat-maligned RJ45 plug doesn't get a chance to trash the connector on my laptop's motherboard.

So why not design a similarly-robust connector for Ethernet?

I mean: We'd probably want a new, standard connector which is totally incompatible with RJ45 for desktop Ethernet devices, anyway. That way there can be low-voltage DC available on the connector, without causing any conflicts with the 48VDC of IEEE 802.3af Power-over-Ethernet*.

This would offer all of the advantages of Ethernet like long inexpensive cables, lots of connected devices, sane and well-understood spanning of multiple network segments, and universal compatibility. It would also offer all of the advantages of USB, like functional plug-and-play, bus power, and low implementation cost.

It might make sense to make a new protocol layer for the thing, something easy (cheap) to implement that has no reliance on things like DHCP. But then, given the ever-decreasing cost of silicon, it might also be just as well to use good ol' IP for everything anyway. (Routable protocols for a desktop mouse or a desktop hard drive? It sounds silly and far-reaching, but seriously: Why not?)

*: I have no interest at all in having a 48-volt inverter in my laptop, for example. There's enough potential there to be painful if things aren't just-so, and I prefer coffee. Better to keep the thing running at something sane like 5VDC. :)

Re:Honest Question

[adolf]

I believe that the SCSI module in linux handles firewire and USB, so from that standpoint it looks like it's a start towards universal communications, except for Ethernet. (Even though old SCSI is nothing like serial: it's the the ultimate expression of parallel communications, with some similarities to the old HP/GPIB parallel communication standard that's still used in for test communication but used to be a hard drive standard.) I have no idea what Windows does.

Linux supports iSCSI, which is a standard for issuing SCSI commands over IP (and obviously by extension, Ethernet). The protocol itself is nearly half a decade old. Microsoft operating systems as far back as Windows 2000 allegedly support it, and Vista offers it as a bullet-point feature.

FWIW.

Re:Honest Question

[Anonymous Coward]

I have wondered about that too, and I suspect that there is no technical reason,
just business and profit reasons. (Also, why, when protocols like RS-232, RS-422,
and RS-485 can be explained in three pages, why is the USB protocol so
unnecessarily complicated?)

Also, I discovered recently, USB is not really "universal" either.
I plugged a new USB 2.0 camera into my computer's USB port, only to find out
that the camera won't work with a USB 1.0 port! What a lousy protocol. (and
too bad, the connectors are really well designed.)

I guess the solution is to just not buy poorly designed products.

Re:Honest Question

[LlamaDragon]

I'm no expert on these things, so I won't argue against your basic premis, but I do have a couple issues with your post (but I'm days late commenting, probably too late to be noticed, but I'll try...)

Even though there is a PoE (Power over Ethernet) standard, most choose not to implement it by default. The fact that people choose not to use a technology doesn't explain why we can't/shouldn't/don't. Using the same logic on a different argument (because I'm hungry), I could say, "Even though most people have a kitchen in their house, they choose not to use it and eat fast food instead." Obviously, that's not a good argument for fast food.

USB uses four wires, ethernet twice as many. Not necessarily - The most common implementation at 10 or 100 megabit uses only 4 wires (pins 1,2,3,6). This leaves the other 4 for phone or PoE or whatever. Gigabit over copper, however, does require all 8 wires, but you can still run PoE over it (more info here).

Hope they do a better job on compatibility

[dpbsmith]

I do not look forward to the replacement of what was starting to be a reliable, ubiquitous standard that "satisficed" with a New! Improved! version that shows no signs of actually achieving significantly higher throughput with current devices. Why does USB have to compete with SATA? Why can't USB just be USB?

I've been seriously disappointed with the number of times I've interconnected USB 1.1 and USB 2.0 devices and had them almost work, only to encounter various strangeness and glitches. I don't know who's to blame... whether it's a fault in the standard or in vendors' faulty implementations... and life's too short to care, because know who's to blame wouldn't do much to help solve the problem.

On the whole, I blame the standard, because these days standards are so incredibly huge, bloated, and complex that it is extremely unlikely that anyone actually implements it fully correctly.

With today's sloppy practices of testing to the market ("Let's try it with the most popular devices, or the ones which are most important to our business") instead of testing to the standard, the result is all sorts of opportunities to build devices that comply with the standard but do things just a little differently than the most popular devices... and have them not work even though they "should."

A typical example was an IOmega external CD burner I bought once for a USB 1.1 Mac. (I chose it because it was $30 cheaper than a FireWire model, I wanted both PC and Mac present and future compatibility), and I didn't really care about speed. The drive actually burned perfect CDs, but it always claimed erroneously that an error had occurred. But how could a sane person rely on that? I returned it, bought a different USB 2.0 external CD burner from a different vendor... and encountered exactly the same problem.

I've also seen various glitches and strangenesses trying to use USB 1.1 thumb drives in USB 2.0 CPUs and vice versa.

Re:Hope they do a better job on compatibility

[aneviltrend]

I've been seriously disappointed with the number of times I've interconnected USB 1.1 and USB 2.0 devices and had them almost work, only to encounter various strangeness and glitches. I don't know who's to blame... whether it's a fault in the standard or in vendors' faulty implementations... and life's too short to care, because know who's to blame wouldn't do much to help solve the problem. It's not the specification that is to blame. I've spent the time implementing a USB 2.0 host controller, and I can say that the specification is definitely able to handle the mix of USB 1.1 and USB 2.0 devices on it.

the result is all sorts of opportunities to build devices that comply with the standard but do things just a little differently than the most popular devices... and have them not work even though they "should."

This is exactly where the problem is: with the device vendors. There are a lot of choices for USB hardware, and you get what you pay for. With the 600+ page spec that USB 2.0 comes with, it's easy to trip up and miss a corner case that the spec defines. But bear in mind that the spec does define that corner case, and what to do when you reach that. It's up to the vendor to actually implement that.

USB 3.0 will have its own fair share of sloppy vendors that don't spend the time to implement the spec fully, and that will cause a lot of headaches for end users. As the products mature, however, we should see that USB 3.0 will be a powerful protocol. I see a lot of complaints on this article about the 'actual' vs. 'projected' speeds for USB. These posters need to understand a few things:

When the spec states that information flows across the bus at 480 Mbits/s, that means that the frequency at which the bits toggle on the line is 480 Mbits/s. For data to be sent across a USB line, the host must first initiate the transfer with a 'token' packet, which is about 2.5 bytes. There has to be about half a byte of 'delay' on the line after that, at which point the sync for the data packet can be sent, which is about a byte. Only at this point can the actual data be transmitted. But it's not even that easy, because the data is NRZI-encoded and bit-stuffed as well. This is why there's a lot of CPU overhead with USB data transfers - the CPU has to decode the data packet. After the data packet is sent, there's at least 3 more bytes of stuff that needs to happen before the entire transfer is complete.

There's even a limit on the size of the data packet: in USB 2.0, for a bulk transfer (i.e. when you're copying a file to/from your external drive) the maximum data that can be transmitted in one data packet is 512 bytes. And there's no guarantee that you'll even be granted a data transfer at a constant rate - it's completely up to the host controller how often it'll ask the device for a data transfer. If there's a lot of devices contending for the bus, your up/download rates will suffer.

There's a lot of necessary overhead inherent in an asynchronous communication protocol, especially one that grants you the ability to put multiple devices on the same bus. End users need to understand that data does not magically "pop" from one end of the cable to the other, and wherever there are different people implementing the same spec there will be slight differences in how the devices end up working.

Re:Hope they do a better job on compatibility

[macslut]

Connecting a CD burner to a USB 1.1 Mac + Resulting error = Your fault "I do not look forward to the replacement of what was starting to be a reliable, ubiquitous standard that "satisficed" with a New! Improved! version" Personally, I thought the upgrade to 2.0 was a really great thing and believe 3.0 could offer quite a lot for those of us who may be wanting to transfer terabytes on a regular basis. The idea of a one-port standard for everything from keyboard to drives is pretty compelling. I just hope as a previous poster stated, that it gains systematic efficiency, making it at least as desirable as FireWire in that regard.

Re:Hope they do a better job on compatibility

[Anonymous Coward]

But does the standard really need to be that long? How can you blame somebody for missing something in a 600 page document. There are other standards out there that handle complex transport problems but do so in much simpler ways, which eliminates confusion and leads to better implementations.

Easier to use?

[Wiseman1024]

"This should make USB hard drives easier and faster to use."

And exactly how is more bandwidth making things easier to use? This statement is like crap straight out of a business magazine.

Re:Easier to use?

[aadvancedGIR]

Moreover, you'll have to make sure you have one of those new cable. Getting rid of the power cable is good, but it is something we already have on some USB2.0 external HDD.

Don't let the marketing guy name it this time.

[JoeD]

I can never remember which one is faster, "Hi Speed" or "Full Speed".

we're probably going to wind up with yet another ambiguous name like "Extreme Speed" or "Max Speed".

Just call it USB 3.0 and be done with it.

Re:Don't let the marketing guy name it this time.

[Anonymous Coward]

Just call it "USB Max Power!"

Max Power never apologizes for dropped packets or bus conflicts!

                  - Rembrandt Q. Einstein

Re:Don't let the marketing guy name it this time.

[shellbeach]

I can never remember which one is faster, "Hi Speed" or "Full Speed". As I was explaining to someone the other day, you've got the alternative of "Full Speed", which isn't, or "High Speed", which is full speed.

Personally, I was hoping they'd call this new standard "Full High Speed", and then the next standard "High Full Speed". After that can come "Highly High Full Speed", "Fully High Full Speed" and "Speedy Full High Speed".

I'm sure it makes more sense if you're high at the time ...

Re:Don't let the marketing guy name it this time.

[mjwise]

I'm sure it makes more sense if you're high at the time ...

Well, I am full...is that good enough?

Re:Don't let the marketing guy name it this time.

[SoulRider]

How about, "Ludicrous Speed"

Won't work till 2015...

[flyingfsck]

It's OK, in typical USB standards fashion, it won't work properly for the first 5 years, so you have until (at least) 2015 to buy a new printer.

What's worse is..

[bogie]

Working on computers that put the usb ports 1" from the ground. Or even better hide them between the network port and all of the other ports. Real easy to access either port location when the computer is shoved where ever it fits and is out of the way. How about moving the ports to the top of the computers and have them lit up all them time so that you can find them instantly? Would that be so freaking hard? You can even put a little cover over the ports with your oem logo on it or something.

Best usb port design? 5 year old Dells which have the usb ports at the front bottom with the ports ANGLED UPWARDS. It's lotsa fun trying to fit a thumb drive into those ports. I'd love to punch the guy in the face who came up with that design.

Re:What's worse is..

[eh2o]

Oh man... do let me know if you ever find that guy, I'd like to kick him in the stomach after you punch him!

Re:What's worse is..

[Guysmiley777]

Cooler Master figured it out. I love this case:

Centurion 532

Re:What's worse is..

[karnal]

I am SO buying that case for my next build. Thanks!

What mod-with-a-vengeance called this "flamebait"?

[Anonymous Coward]

The guy describes a repeated technical problem, in detail, and is modded down not just once but several times as flamebait?

Looks like you pissed someone off...

Re:speed over fiber vs speed of electrons over cop

[Billly Gates]

I am not an engineer.

But the faster you push data on copper the more vulnerable to distortion and corruption the data becomes. Wires act like antennas and absorb em radiation from the computer and other sources. This is why gigabit ethernet has a very short distance the cable can cover vs 100mbps cables.

Light doesn't suffer from this problem and thus can handle faster data.

Re:Eat into firewire not likley

[J Isaksson]

I suggest you look into isochronous transfers on USB, which indeed can guarantee a minimum bandwidth.

opto AND wire?

[TheGratefulNet]

why both?

sounds like its expensive to make, cant be cut to custom lengths (easily), can't be wrapped in a tight radium (fiber optic never should be), and there are already good standards for high speed (its called 'gig ethernet' and 'firewire' and 'sata').

add more differential pairs if you NEED to increase speed but hell, do NOT add mixed fiber and copper. that's the dumbest thing I've heard since I posted a joke article about differential mode fiber optic audiophile cables (an apr-1 joke).

speaking of bad standards, the worst thing I've seen come from connectivity standards is that HORRIBLE thing called sata data and power. the power connector offers NOTHING useful over the 4 pin molex and in fact its less stable since it holds 'in' much less than the very tight 4 pin molex power plugs. the sata data cable is too stiff, the cable wants to form its own shape and NOT be a slave to where the connector is clicked in. you first have to get the wires 'molded' to the shape of the case and path they'll take from board to drive, then you have to pre-stress them so they won't follow their own direction and pull out or work themselves out. a stran relief is STRONGLY needed but never designed in ;( now they finally realize they fucked up and they added locking latches to the sata data connector. too little too late. the wire is TOO STIFF for the junky connector you guys used. also its not keyed well enough so you have try it one way, waste time and effort and then if it doesn't fit try it the other way 180 rotated.

what a joke - the industry can't seem to design a decent mechanical connector! ;( it really is sad. I wonder if this new hybrid opto/cable CRAP IDEA is going to have badly made connectors and overly stiff wire as well?

Re:opto AND wire?

[Anonymous Coward]

the SATA power connector has one useful feature.. it supports hot-plugging of SATA drives..., traditional molex doesn't

Re:opto AND wire?

[Anonymous Coward]

Wow, that's amazing. Almost everything you said is WRONG.

What happened?

[allaunjsilverfox2]

I thought that WUSB was supposed to replace USB 2.0? Is there going to be two different standards now? And is there going to be a interoperability between WUSB 1.0 and USB 3.0? That would kinda interesting to see done.

tenfold or 10 times

[Anonymous Coward]

If it is currently 480mbps, a tenfold increase would take it to 480*2^10 = 480 * 1024 =~ 480gbps.

Re:tenfold or 10 times

[TechnicolourSquirrel]

'If it is currently 480mbps, a tenfold increase would take it to 480*2^10 = 480 * 1024 =~ 480gbps.' Sorry, ten-fold is a synonym for 10 times, not for 10 doublings. What you've just done is a thousand-fold increase.

Re:speed over fiber vs speed of electrons over cop

[MSDos-486]

Yes there is 10Gig over copper. However The current standard 10GBaseCX is only over short distances, they are working on using UTP but its Cat7 UTP so it doesn't work with current cable.

typo

[TheGratefulNet]

sed /radium/radius/

(too early to type well..)

Re:typo

[Anonymous Coward]

I think you might have some problems getting meaningful amounts of current across fibre.

Re:typo

[TheGratefulNet]

WHY have to bundle them if you need power?

I agree - sometimes its useful to have power and signal in the same jacket but then, why go optical?

has wire been 'used up' yet? I hardly think so! shades (heh) of the old atm vs ethernet wars when opto atm was touted as running faster 'than ethernet ever would'. yet it was a lie and today, who the hell runs optical cable unless they have to? wire works JUST FINE at the same speeds. you might need decent a/d converters to extract the signal but same kind of cost as opto blocks (and opto just a PITA to deal with!).

Only 4.8Gbps?

[Doc Ruby]

We've already got 10Gbps optical ethernet for much longer distances than USB will support, at only $40 for a PCI-X adapter. I wonder why we don't have higher speeds for shorter distances like USB peripherals. I understand optical makes the distance limits less important between a few mm and several hundred meters, but why not at least a 10Gbps USB already? If it's because ethernet is parallel, there's also 10GbE over twisted copper pair, so why not a 10Gbps USB?

What about PoE

[Anonymous Coward]

Ethernet is all signal, no power. You couldn't charge your iPod or iPhone with your sync cable, nor would your mouse or keyboard work without batteries.

That would be "Power over Ethernet".

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

USB is simplex, FireWire is Duplex

[IvyKing]

In high speed mode, the same pair of wires is used for transmit and receive - FireWire has separate pairs.

My concern is

[iLoveYoyo]

while people talk about backward compatibility, i am concerned with "forward" compatibility. Can a USB 3.0 flashdrive can be inserted into a USB 2.0 port on most laptops and PCs nowadays???

You win a toaster!

[Anonymous Coward]

This is hilarious. You're the only one that got that right.

USB much slower than disk drives

[Anonymous Coward]

Interesting that nobody on this thread seems to know what real world USB 2.0 performance is.

Drives that transfer at 80 MB/second via SATA or ATA only manage 20-25 MB/sec under USB (under Linux and XP). It takes 5+ hours to format a 500GB USB drive or 11+ hours to copy a terabyte.

Cost?

[blhack]

So best buy is going to charge, what, $400 a cable for a 3 footer now?

"But you don't understand! Data on this cable moves at the speed of LIGHT!!! THIS FIBER OPTIC CABLE IS GOLD PLATED!"

Re:speed over fiber vs speed of electrons over cop

[Anonymous Coward]

Two things:
1 gibbit ehternet max distance is the same as 100 ethernet.

2. The reason why fiber is used for long haul is minly about energy loss, not about distortion(but yes distortion occurs, but thus also
happens on fiber)

Wrong

[RingDev]

USB 3 is 4.8 Gigabits per second.
That's .6 Gigabytes per second. .6GB/sec = 600MB/s. .9TB = 900,000MB.
900,000/600 = 1500 seconds

Assuming the drive itself is empty, formated and has a sequential write speed to keep up, you are looking at 25 minutes to fill a .9TB drive.

-Rick

Re:Wrong

[bored]

You forgot to subtract protocol and bit encoding overhead. In reality there is probably about 480MB/sec of bandwith on a 4.8Gbit link (see 8B/10B encoding) then you have to subtract protocol overhead which for USB probably will burn 10-20% of that. So the end result is probaby about 400MB/sec.

Useless *almost*

[MBHkewl]

As SATA2 disks give a throughput of 70 MB/s, the 4.8Gbps throughput is kind of useless, unless the USB connection was used to connect multiple computers, forming a LAN.

Re:speed over fiber vs speed of electrons over cop

[pixelkiller]

What they need to do is put a cpu on that bus so that it will take away from the cpu overhead / Usage.
Sorta like what there doing with video cards having there own ram and cpu. last thing you need is a 3.8 Gbs
drive using valueble Clock ticks / Cpu Time.

Not so impressive

[damieng]

...given that USB 2.0 (480MBits) was 40 times faster than USB 1.x (12MBits) ;-)

[)amien

0.9 TB / 4.8 Gb/s = 1500 seconds

[Anonymous Coward]

WTF, a chain of "Exactly" and "Indeed", and no one realizing that 1.5 seconds is wrong by a factor of 1000?

Re:0.9 TB / 4.8 Gb/s = 1500 seconds

[atomicthumbs]

Indeed. (they probably mistook 0.90 for 0.09)

Re:0.9 TB / 4.8 Gb/s = 1500 seconds

[thegsusfreek]

They probably work at Verizon.

Re:0.9 TB / 4.8 Gb/s = 1500 seconds

[JazzLad]

New 9GB Drive?

Re:0.9 TB / 4.8 Gb/s = 1500 seconds

[thecountryofmike]

Hmmm. No.

0.9TB = 900 GB

at 4.8GB/s

we have:

900/4.8 = 187.5 seconds (~3 minutes). And off by 2 orders of magnitude.

Re:0.9 TB / 4.8 Gb/s = 1500 seconds

[Anonymous Coward]

you're forgetting the difference between bits and bytes

0.9TB = 900 GB
  at 4.8GBits/s
  we have:
  900/(4.8/8) = 187.5*8 seconds = 1500 seconds

Re:0.9 TB / 4.8 Gb/s = 1500 seconds

[GeePrime]

Ok, here's the deal.

it's Gb/s, not GB/s, that's gigabit, not gigabyte.

If he is referring to a disc manufacturer terabyte, then the answer lies here:
http://www.google.ca/search?hl=en&q=900000000000+bytes+%2F+4.8Gbps&btnG=Search&meta=

(about 1 397 seconds)

If he is referring to an actual terabyte, then the answer lies here:
http://www.google.ca/search?hl=en&q=.9TB+%2F+4.8Gbps&btnG=Search&meta=
(about 1 536 seconds)

We really need to change our conventions, little b for bit and big B for byte are too similar, and cause confusion. As well, the fact that kilo/mega/giga/etc...byte mean different things to different people is just terrible. the GiB solution proposed (Gebibyte, or binary gigabyte) solves the problem, and I believe is accepted by IEEE.

Re:0.9 TB / 4.8 Gb/s = 1500 seconds

[thecountryofmike]

Hee hee. My bad.

I realized it too late. Damn bit vs. byte! Why don't they just quote transfer rates in terms of BYTES?!?! I know, I know, then the numbers wouldn't be as impressive...

*hangs head in shame*

Re:0.9 TB / 4.8 Gb/s = 1500 seconds

[Workaphobia]

The binary alternatives to the SI prefixes - Gi, Mi, etc., - make me cry. It's like we're *trying* to obfuscate the pronunciation of technical terms. It's bad enough we have to deal with "www". On a separate note, I dare you to walk into a computer store or any other public place, and engage in a casual loud conversation that includes the word "Mebibyte", and see how many looks you get.

Re:0.9 TB / 4.8 Gb/s = 1500 seconds

[GeePrime]

If I post it on YouTube, will you watch? (I'm crazy enough to do this.....)

on a serious note - what do you propose as a solution to the problem? We have same names for different things, how else can we solve it? It's easy to point out the faults with an issue, but it's hard when you have to solve it. a kilobyte, as we know it, uses the SI prefix, but it truly isn't. a SI kilobyte, by definition, would be 1000 bytes, not 1024. We make engineers cry when we use these prefixes improperly. We need different terms for different things is the real problem, and maybe you are right, the binary prefixes are terrible. but if that's the case, we need a better solution.

Re:speed over fiber vs speed of electrons over cop

[Shotgun]

GE is limited due to the speed of light and the way the ethernet protocol works. A sender has to stop sending if it senses someone else talking on the same line. In order to do this, it has to detect the collision before it finishes sending. If the line is to long, a sender at each end will be able to get an entire packet out before being able to sense the first bits from the other end. Ugly things happen then. Google "CDMA-CS" if you really want to know more about what limits the length of ethernet.

EM interference is handled by the twisted pair. A pulse of EM energy will cut across the signal and ground wire at the same time. The reciever senses the difference in the voltage levels across the pair, so if you effect both at the same time, the reciever doesn't know (or care).

Fiber has it's own host of problems, but for these short distance and relatively low data rates (for optics) they can use lossy plastic cables with 1/8" headers and just pump LED power to make up for the loss.

Reaction?

[j-min]

It seems when given the choice to embrace a technological advance that may not be useful today but perhaps tomorrow or to give a pedantic rant about how much they know about hard drives and converting bytes to bits, many has chosen the latter.

Ok fine, so hard drives have physical limitations keeping them from using USB3 to its full throughput, but if 5 years down the road there's a new kind of hard drive, I imagine there'd be the same group complaining that USB3 isn't fast enough, blah blah blah who needs these faster hard drives when USB3 is the bottleneck.

What goes at 3Mbit/sec?

[heroine]

Still waiting for a hard drive which reads over 20MB/sec in the real world.

Re:What goes at 3Mbit/sec?

[|Cozmo|]

You must have some old crappy drives. My disks can do upwards of 80mbytes/sec.

System overhead is an issue!!

[SuperKendall]

System overhead on a dual core or quad core and above CPU is not going to be an issue.

System overhead even on modern dual-core CPU's is ALREADY noticeable, with USB 2.0 drives which is why I use only firewire enclosures if I have a choice. You are saying that overhead from a spec 100 times faster is not going to also be noticeable?

Re:System overhead is an issue!!

[Anonymous Coward]

High CPU utilization == high power consumption too.

Improved power model

[caywen]

I'd like to see USB 3 improve on its power model. First, I want USB 3 to have potential of providing more juice to peripherals like 3.5" external drives, external DVD burners, desktop speakers, printers, and even (perhaps small) external displays. Currently, USB 2 can barely power a 2.5" external drive. I think of of the goals should be to reduce the number of bulky AC adapters dangerously hanging out of power strips.

Title is not accurate

[master811]

Only the specification is coming in 2008, actual USB 3.0 you won't see until 2009/2010

Backwards Compatible?

[nurb432]

Or do we have to ditch all our existing USB devices?

MathTastic...

[tempest69]

ncredible and grand news if it's on target, and doesn't dissolve into vapor. Cue the Media Copying Discussions. (Someone fast on their math: How long would that take to copy a new 0.90 Terabyte drive?)

Current USB ~= 60 MB/Sec Theoretical

USB3 ~= 600 MB/Sec Theoretical max,

1000 sec ->600 G *1.5 == 1500 sec for 900 G

1500 Sec/(3600 Sec/hour) um 25 minutes

In real USB drives run about 25-28 MB/sec under half the theoretical rate. 500 mins ~ 8+ hours

Modern Hard Drives can pull 60-75 MB/sec sustained. 250 mins- 200 mins (respectively) or 3 to 4.5 hours.. (plus a nice fudge factor, as the inside tracks have a slower transfer rate)

Storm

Gigabits != Gigabytes!

[kcbanner]

"lol 4.8 gig a second lol"
This is what I'm seeing. Most people fail to see that a byte is indeed 8 bits! Miraculous indeed. Guess why they say things in gigabits/s. Its a bigger number, therefore the masses get attracted to it. Huzzah!

Re:speed over fiber vs speed of electrons over cop

[bored]

This is why you should buy a good switch (instead of a hub). Twisted pair is full duplex and a good switch will simply buffer anything sent from multiple sources at the same time. The carrier sense aspects of ethernet are basically dead, if the swtich buffer fills up for a particular destination it can throttle the source using the pause aspects of modern ethernet.

4.8Gbit USB ~= 400MB/sec != 600MB/sec

[bored]

Ok given a 4.8Gbit link the real transfer speed is probably about 480MB/sec after 8b/10b encoding, then given a 10-20% hit for the USB protocol, your probably talking about 400MB/sec, not the 600MB/sec other people are assuming.

someone please explain point of USB2 even

[2ms]

From the perspective of someone into digital music recording and production, USB 2 seems like a terrible thing -- it sends data in clumps with high latency not particularly fast, whereas Firewire 800 is insanely fast with no latency and no problems with "smoothness", but computers have all come with USB2 instead of Firewire ever since 2.0 came out.

Could someone explain what the justification behind the vast majority of computers coming with USB 2 rather than Firewire? I realize the reason for it is prblem Intel promotion etc. But could someone tell me any reasons why it isn't just completely inferior to Firewire for the consumer?

USB3 strikes me as just an extension of this sad thing of computers not using Firewre because USB exists.

this is not your father's fiber

[statemachine]

I've seen the current-gen fiber extremely abused and it still works. Ninety-degree turns, tight coils, stepped on constantly, no slack... (No, it's not ideal, and these conditions should be avoided.) But, for a short-haul connection of a few meters, you will likely not see any problems unless you snap the fiber in half.

Re:Plug Shape (SCSI was worse)

[Cochonou]

Incidentally, Firewire is actually a SCSI-3 standard.

My questions

[Kaenneth]

1) Will I be able to charge my phone on it?

2) Is just the raw data speed increasing, or are other limits being increased?, I heard that USB drives currently have a 2 terabyte limit. (Drobo)

3) will humping dog USB toys be prevented?

oh noes!

[kaizokuace]

what ever happened to firewire! I thought that sustained transfer was much nicer than burst. Oh well. Also you didnt always put the plug in upside down first like with usb.

(Non-)Obligatory Back to the Future Reference

[JazzCrazed]

Four...point...eight...JIGGABITS???

Great, as long as they change the plug

[kalislashdot]

I dont care, just change the plug. I hate it. Who ever designed the rectangle plug wanted to inflict torture on us all. Firewire has a better plug, the other USB plug id better, the B side, the side in the printer. Why or why did they make to look the same so I have to fiddle to get it in.

I prefer FireWire

[OrangeTide]

It's a simpler model, has less load on the host system, and devices can peer. There are even variants of FireWire that can already go 3.2Gbps over an optical connection. Also you can pull a pretty significant power load from a FireWire port (typically 5W up to 30W with the right host). As an occasional device driver author, I know I prefer implementing FW drivers versus USB host stack drivers(yes, I've done this), and the same goes for the class drivers.

Not that my preference matters to the industry. It seems PCs are hell bent on making 1394 an optional feature on motherboards, and limiting the technology to the A/V realm. It might be because you have to have USB, because FireWire is no good for a mouse and keyboard. And you can do external drives with USB and eSATA. FireWire, although superior in many ways, is not a bare minimum requirement. USB is a bare minimum requirement, so extending its complex (and terrible) design will lead to wide adoption by the industry.

That being said...

[Ayanami Rei]

The bulk transfer mode should have allowed endpoints to send unsolicited packets in the open timeslots, however, rather than using a one-shot "okay let's see if anyone wants to transfer" approach. All that's needed is a support for a HUB-supplied NAK when an existing transfer from a different endpoint is occuring -- and a nice HUB might even store and forward.

Could this be used for video?

[Fian]

Anyone know what bandwidth would be needed to run your monitor over USB? It would be nice if you could buy a PC with only USB ports (no serial, parallel, PS/2 etc) it would make setup and support so much easier.

Switches, not routers

[lennier]

If you want to be technical. Switches are Layer 2 Ethernet devices, routers are Layer 3 IP devices. Ethernet is a single-segment link-local nonroutable protocol and cares nothing about TCP/IP.

And if you want to be even more technical about it, it is becoming increasingly difficult to find non-switched (old-school CSMA/CD) Ethernet nowadays in the pervasive 1 Gigabit world. Even cheap four-port home Ethernet 100Mbps 'hubs' are in fact switched. So it's not your granddad's Ethernet anymore, and many of the efficiency arguments about CSMA/CD don't apply.

Re:Switches, not routers

[gillbates]

USB allows multiple hosts on the same bus without collisions. Ethernet requires only one host on the same physical "bus" to avoid collisions. To get switched ethernet, you have to use a router, which requires additional cost.

It doesn't do that now, so...

[Ayanami Rei]

nt

The Answer

[joe_bruin]

Do you see the USB icon on one side of your cable? The one that looks like one path forking into three? That one always faces UP. So on the ports on the front of your computer, it will be facing up when you plug it in correctly. If the ports in the back are vertical, it will be UP in relation to how your motherboard is seated (which is pretty much the same on all standard desktops). I hope this clarifies any confusion you have.

Cool!

[real gumby]

I already have a bunch of cool USB-powered doohickies. Now I can get USB 3-powered doohickies. Lesse,..USB ports are supposed to put out 500 mA at 5V...hmm, cool, that' about 400x what a laser pointer is allowed to produce.

Yep, I can't wait: a hub of these things...and some USB 2-controlled mirrors...

inconsequential nitpick

[j1m 5n0w]

I believe that's only true of singlemode fiber. Multimode usually uses LEDs.

Re:Great. Gee. Golly!

[aqk]

Hmmnn.. You don't have much imagination eh?

The USB3 will be attached to a HD farm. Lots of arms and heads.

That is if the HD isn't obsolete by then.
You'll be booting from a built in Flash drive. etc etc...

C'mon guys! Ths is just the beginning! Please one-up me with your wild ideas.

Re:Bottleneck? Windows

[Anonymous Coward]

Windows XP and 98 at least have had the shittiest slowest USB implementations I have seen in any OS. I can tell you with 2 USB enclosures I have right now.. well, I have a 250GB enclosure plugged into a Inspiron 1100 (Celeron M 1.4) that (just running a few tests) got 29-30MB/sec at about 3% CPU utilization. A Sempron 2500+ system with a 1TB enclosure got 34-35MB/sec at 10% utilization. Not great, I'd prefer 55-60MB/sec at 1% utilization.. but anyway. I am running gentoo on these but have seen similar speeds with Debian & Ubuntu, and also FreeBSD, NetBSD and MacOSX.

Re:Plug Shape (SCSI was worse)

[Anonymous Coward]

Sweet jesus... I have to deal with surplus SCSI equipment at my job. They'll ship systems in with no disks, then boxes of random SCSI disks will come in weeks or months later. It's so ridiculous. Since departments hold on to stuff for RIDICULOUS lengths of time we get every type in... We've at least gotten all the below...

          Centronics (the REALLLY old enclosures like Apple 80SC etc..), miniature 50-pin, 68-pin and 80-pin. Don't get me started on SSA. That's the external types.

          Internal -- 50-pin (bit wider than IDE), a miniaturized 50-pin, 68-pin, 80-pin, SCA, plus plenty of slightly non-standard brackets and adaptors to fit the whim of Dell, HP (with Netservers and HP-UX line being totally different), Compaq, IBM, etc. etc. all being a bit different. Again don't get me started on SSA.

          Oh, plus the joy -- "Hey, I think this'll fit!!" "OK... NO wait it's LVD!!!" *bzert*

          For some perverse reason they had regular SCSI, LVD (low voltage differential) and (uncommonly..) HVD (high voltage differential) WITH full ability to mix and match them, except they'll blow each other up when you actually power the system on. (Maybe you'll blow a disk.. maybe you'll blow a controller.. maybe you'll blow both.. it's fun to find out which.) For whatever reason, some vendors just decided it wasn't necessary to label hard drives as LVD or HVD.. once in a while a system vendor decided they didn't have to label the ports either. (Luckily, they usually do.)

          I've had people both come in to get surplus equipment, and people I work with, that absolutely get wood over Ultra160 and Ultra320 for running home servers. Fuck that shit. IDE is slower, but there's just 40-pin ATA, 80-pin ATA and SATA. And if someone gets sloppy and hooks a drive up with a 40-pin cable it'll just slow it down a bit. Foolproof!