superhighspeed usb 3.0

This is a discussion on superhighspeed usb 3.0 within the Tech Board forums, part of the Community Boards category; USB 3.0 speeds data | Tech Blog | FT.com It is five years since the USB 2.0 standard was rolled ...

  1. #1
    Registered User kryptkat's Avatar
    Join Date
    Dec 2002
    Posts
    638

    superhighspeed usb 3.0

    USB 3.0 speeds data | Tech Blog | FT.com

    It is five years since the USB 2.0 standard was rolled out, providing a common high speed data transfer protocol for a swath of CE devices ranging from external hard drives to smartphones. Since then however, read/write speeds have continued to climb and USB 2.0, which has a maximum theoretical throughput of 480Mbps, has become something of a data transfer bottleneck.

    USB 3.0, which can theoretically handle up to 5Gbps is designed to eliminate this bottleneck delivering transfer speeds up to 10 times faster than USB 2.0 devices. To put that in perspective, USB 3.0 is fast enough to download 8Gb of photos from a digital camera in just 20 seconds or download a 27Gb HD movie is just over 1 minute.

    they said that usb 3.0 is suppose to come out on windows 7 machines. and is usb 2.0 backward compatible. have you seen the usb 3.0 ? they said it uses 500ma on usb 2.0 and 900ma with usb 3.0. mostly for msd that could improve performance greatly. it could also improve usb wifi adapter speeds too. due to be available to consumers later this year.

  2. #2
    Devil's Advocate SlyMaelstrom's Avatar
    Join Date
    May 2004
    Location
    Out of scope
    Posts
    4,074
    There are a number of P55/H55/H57/X58 (Intel's Core iX series) motherboards that have USB 3.0 support already. The pin-out between the two specifications are the same, so a USB 2.0 cable will work fine in a USB 3.0 bus and visa-versa. The difference is the controller on the motherboard. Flash memory sticks will also have to be USB 3.0 specific as the memory controllers on the flash for USB 2.0 aren't designed to handle the throughput, though you might see some gain in putting one into a USB 3.0 bus. I've even seen "USB 3.0 cables" which promise additional throughput, but I'm not certain they're necessary. Perhaps to get the maximum speed of USB 3.0 you might need the specific cable, but I'm pretty sure any old USB cable should handle it just fine.
    Sent from my iPad®

  3. #3
    Internet Superhero
    Join Date
    Sep 2006
    Location
    Denmark
    Posts
    964
    Quote Originally Posted by SlyMaelstrom View Post
    I've even seen "USB 3.0 cables" which promise additional throughput, but I'm not certain they're necessary. Perhaps to get the maximum speed of USB 3.0 you might need the specific cable, but I'm pretty sure any old USB cable should handle it just fine.
    Sounds like those very expensive gold-plated HDMI cables which, according to the manufacturers are necessary for good quality HD, but which are a complete waste of money according to everyone else :-)
    How I need a drink, alcoholic in nature, after the heavy lectures involving quantum mechanics.

  4. #4
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    USB 2.0 cables may not be specified (rated) to the level required by USB 3.0 speeds in terms of bandwidth (analog bandwidth), crosstalk, etc.

    So it's possible that while USB 2.0 cables will USUALLY work, it's really undefined behaviour.

    It's like running gigabit network on Cat5 cable (instead of Cat5e or Cat6). It works 95% of the time.

  5. #5
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    Actually, according to Wikipedia
    When operating in SuperSpeed mode, full-duplex signaling occurs over two differential pairs separate from the non-SuperSpeed differential pair. This results in USB 3.0 cables containing two wires for power and ground, two wires for non-SuperSpeed data, and four wires for SuperSpeed data, and a shield that was not required in previous specifications.
    So if you use USB 2.0 cables, you can only go at USB 2.0 speed (high-speed), because it's missing 4 wires.

  6. #6
    Registered User kryptkat's Avatar
    Join Date
    Dec 2002
    Posts
    638
    i was reading through the specs and it said usb 3.0 uses
    "double simplex" instead of full duplex.

    the difference would be ?

  7. #7
    (?<!re)tired Mario F.'s Avatar
    Join Date
    May 2006
    Location
    Portugal
    Posts
    7,510
    Probably that communications are unidirectional, as opposed to duplex communications that are bi-directional. 'double' because probably allows for two channels to send data at the same time.

    It's twice a half-duplex.
    The programmer’s wife tells him: “Run to the store and pick up a loaf of bread. If they have eggs, get a dozen.”
    The programmer comes home with 12 loaves of bread.


    Originally Posted by brewbuck:
    Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.

  8. #8
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    USB 2.0 was half-duplex. Both directions share the same pair of wires (differential pair, meaning the difference in voltage between them represents data), so only one side can send at a time.

    USB 3.0 introduced 4 more wires (2 more pairs), so it can do full-duplex (both sides can transmit at the same time).

  9. #9
    (?<!re)tired Mario F.'s Avatar
    Join Date
    May 2006
    Location
    Portugal
    Posts
    7,510
    Yeah telecommunications are weird that way.

    But probably they call it double simplex because it only emulating full duplex. It's probably two pairs that send data at the same frequency, but the controller can switch one of the pairs to invert direction.

    Confusing... but I'm just making it up.
    The programmer’s wife tells him: “Run to the store and pick up a loaf of bread. If they have eggs, get a dozen.”
    The programmer comes home with 12 loaves of bread.


    Originally Posted by brewbuck:
    Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.

  10. #10
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    I think what you are describing is half duplex.

    In full-duplex, data never changes direction. One pair is always host to device, and another pair is always device to host.

  11. #11
    Registered User kryptkat's Avatar
    Join Date
    Dec 2002
    Posts
    638
    i understand what simplex is and i understand what duplex is. simplex like a transceiver while duplex is like a phone. one line out one line in. but where is the advantage speed wise ? if you have 10mb on both sides <two computer> to send to each other you can do it in have the time per file but over all in eighter mode it still takes the same amount of time to xfer 20mb.

    if i was going to design usb 4.0 i would first make the connector 4x wide with 64 lines instead of two. thinking why not 4 or 8 or 64 ?

    also i would make for the above usb 4.0 external plug in cards such as an additional graphic processor card. why not you already have msd why not add a processor or a few to make a graphics card ? i would also make usb 4.0 plug in multi core processor cards ? why limit your multi core processor machine to just one processor chip ? how about extra math coprocessors plugged in to the usb 4.0 ? how about graphic accelerator card usb 4.0 plug in instead of just another graphic card ? the difference being extra math coprocessors on it. or a multi tuner tv card say with 8 or 9 or more tv atsc tuners and one or two stsc tunners ? or a atsc multi tuner with a graphic accelerator card to plug in to the usb 4.0 ? salivating now .... meow.

  12. #12
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    Where on the page does it say simplex? I don't see it on there at all?

    It seems to me like it's just half-duplex (1.1, 2.0) vs full-duplex (3.0).

    If you have half-duplex, assuming you can do 1MB/s, and you have 10MB each side to send, it will take 20 seconds (only one side can send at a time).

    If you have full-duplex, it would only take 10 (both sides can send at the same time).

    if i was going to design usb 4.0 i would first make the connector 4x wide with 64 lines instead of two. thinking why not 4 or 8 or 64 ?
    That is the debate between parallel buses and serial buses. USB (universal SERIAL bus) is a... well, serial bus. Meaning bits are sent one by one, instead of all at the same time in a parallel bus. In the last few years, almost all parallel links in computers (PCI, IDE, parallel port, SCSI) are being replaced by serial links (PCI-E, Serial ATA, USB, FireWire).

    Logically, parallel buses are much faster, because you are sending multiple bits at the same time.

    In reality, not so much, because the crosstalk between the wires will corrupt the signal (conductors will affect each other), and force you to clock the bus at a lower speed. It also severely limits the cable length.

    USB 2.0 is clocked at 480MHz (and to support a 480MHz square wave, the conductors need to support at least a few harmonics/multiples above that). You'll never be able to clock a parallel link anywhere near that fast. So in the end, serial is still faster.

    More info: http://en.wikibooks.org/wiki/Communi...llel_vs_Serial

  13. #13
    Registered User kryptkat's Avatar
    Join Date
    Dec 2002
    Posts
    638
    in the " usb 3.0 (11132008) final . pdf " under 3.1.3 usb 3.0 system configuration

    4.3 super speed protocol over view

    4.3.1 difference from usb 2.0 under dual simplex unicast buss

    4.3.1.1 comparing usb 2.0 and super speed transactions

    5.3.1.1 usb standard a-connector

    it is under dual simplex instead of double simplex meow bad.

    i am serious about the design that i set forth with the 64 lines design using the superspeed technology. i am very serious about the devices that could be produce to improve computers by just plugging them in as an external device. as you would the msd. changing the serial bus to parallel bus was dry humor. you missed the second subtle humor <grinning like the cheshire cat>

  14. #14
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    Guess I'm good at missing humour.

  15. #15
    (?<!re)tired Mario F.'s Avatar
    Join Date
    May 2006
    Location
    Portugal
    Posts
    7,510
    Looking at that PDF it does say it implements dual-simplex unicast. It's probably a strict adherence to the ANSI.

    In 3.1.1.1 it describes the physical interface:

    - 1 power pair
    - 1 twisted USB 2.0 data pair
    - 1 twisted receive data pair
    - 1 twisted transmit data pair
    The programmer’s wife tells him: “Run to the store and pick up a loaf of bread. If they have eggs, get a dozen.”
    The programmer comes home with 12 loaves of bread.


    Originally Posted by brewbuck:
    Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.

Popular pages Recent additions subscribe to a feed

Similar Threads

  1. question about usb
    By h3ro in forum Tech Board
    Replies: 6
    Last Post: 01-18-2009, 04:02 PM
  2. migrate from .Net 2.0 to .Net 3.0
    By George2 in forum C# Programming
    Replies: 3
    Last Post: 07-25-2007, 04:07 AM
  3. USB Drives???
    By khpuce in forum Tech Board
    Replies: 14
    Last Post: 12-12-2003, 08:01 AM
  4. Usb 2.0
    By ober in forum Tech Board
    Replies: 6
    Last Post: 09-06-2003, 10:14 AM
  5. Printers On USB
    By (TNT) in forum A Brief History of Cprogramming.com
    Replies: 1
    Last Post: 03-10-2002, 10:48 AM

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21