Announcement

Collapse
No announcement yet.

DVI/ATA 100/USB 2.0 pro's and cons

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • DVI/ATA 100/USB 2.0 pro's and cons

    Here's the link to Maximum PC's site with the article regarding the above components.

    The future is bright however if Intel is invlolved, you know performance will be hindered.

    Cheers,
    Elie
    http://www.maximumpc.com/reprint/int...s/revamp2.html

  • #2
    ATA-100 is a transitional standard, and the last ATA drive spec. to use parallel cables. The followup will be SerialATA which is much faster and uses cables akin to a USB cable. Expect them next year.

    Dr. Mordrid

    Comment


    • #3
      Who needs Serial ATA at 1.6Gbps, when you have USB 2.0 at 480MB/sec

      Comment


      • #4
        While the basic SerialATA 1x spec is 1.6 gbps (~200 mB/s) the 3x spec is over 6.0 gbps, or 750 mB/s.

        SerialATA is also backwards compatable with legacy ATA devices by way of a cable adapter. There are also no master/slave setups as all SerialATA devices are "equal" and plug into separate hub sockets on the mainboard.

        In addition ALL the major players are backing SerialATA as the primary storage connector. The heavy hitters; Intel, Microsoft, Maxtor, IBM, Quantum, WD, Seagate, Fujitsu, APT, NEC, Adaptec, Lucent, Vitesse, Promise....you name 'em, they are involved.

        SerialATA support will be in Windows and on mainboards by mid 2001 while there are still no firm commitments for USB2 products from the chip & hardware makers. As such its wide spread appearance in the short term is questionable.

        So far it looks like all show and little go with USB2 seemingly destined for use with external peripherals.

        And then there is FirewireII, which is also faster than USB2.

        Dr. Mordrid


        [This message has been edited by DrMordrid (edited 31 May 2000).]

        Comment


        • #5
          Doc,

          The small "m" stands for "milli", and the BIG one for "MEGA"

          Pertti
          (Sorry, I just couldn't resist )

          Comment


          • #6
            Seriously speaking,

            The biggest benefit of serial ATA would be the capability of connecting more than 4 ATA devices, but then again, the big boys don't seem to see the point, which is not about having more space, but...

            In my first post I was just making a point about how that article happily mixed Bytes with bits, etc.. (the USB2.0 is 480Mb/sec )

            Pertti

            Comment


            • #7
              I think controllers with >4 connectors on the mainboard is entirely possible, even likely. Even if >4 on the board doesn't happen it's very likely that you could add more by way of a PCI card. Remember that Adaptec & Promise are among the partners....

              Dr. Mordrid


              [This message has been edited by DrMordrid (edited 01 June 2000).]

              Comment


              • #8
                It will be interesting to see what the max distance between the controller and drive is going to be??
                paulw

                Comment


                • #9
                  Just wondering,

                  What are they going to do with the HD technology ?
                  At the moment the focus seems to be in improving the interface speed, which is not THE problem.

                  Pertti

                  Comment


                  • #10
                    seems like serial ATA is still as stupid as parallel ATA.... 2 devices per channel? And probably you can't read/write from both at the same time....

                    give me serial SCSI :-)

                    Comment


                    • #11
                      Actually SerialATA does NOT have 2 devices per channel. It has ONE per channel, however a given board might have many devices. The connectors are just arranged on a single "terminal strip".

                      In terms of SCSI it'll be much cheaper and faster than all but the professional SCSI rigs, which none of us can afford.

                      As far as what they are going to do with the technology, in what manner? Speed certainly won't be an issue. Neither will price. Capacity is skyrocketing and access speeds are in single digits.

                      What's left other than transitioning to new technologies in the next decade, which is how long SATA is supposed to last:

                      1. crystal based holographic storage, which IBM is working on as we speak

                      2. flourescent dye storage disk/cards which hit the market next year. 140 gigs per disk, 30 gigs per card.

                      3. atomic resolution storage, which blurs the line between memory and mass storage. Think in terms of at least 10 gigs on a single chip. This one's coming sooner than you'd think....

                      Dr. Mordrid



                      [This message has been edited by DrMordrid (edited 01 June 2000).]

                      Comment


                      • #12
                        Doc,

                        I was referring to the sustained transfer speeds of the HD:s, and the basic idea of the HD:s, which hasn't changed in quite a while.
                        Basically they have just increased the areal density of stored data, and increased the rotational speed of the disks.
                        How much more data can they store on the same area, and when do we have to start wearing helmets to protect ourselves from known flying objects .

                        Ok, the basic technology has matured long time ago, it's cheap due to the production volume, and the transfer speeds are adequate for just about everything, that we do today with our computers.

                        But, what about tomorrow ?

                        My idea of "Tomorrows" computer, would be one without moving parts

                        Pertti

                        (The storage solutions which you added, do sound interesting )



                        [This message has been edited by Pertti (edited 01 June 2000).]

                        Comment

                        Working...
                        X