Follow Slashdot blog updates by subscribing to our blog RSS feed

 


Forgot your password?
Close
typodupeerror
×
Hardware Technology

Serial ATA Drives Mature and Get Faster 189

Posted by timothy from the one-step-at-a-time dept.
MojoDog writes "Serial ATA drives are still as scarce as hen's teeth but what models are trickling out from Seagate and Maxtor, are beginning to look promising. This article and performance analysis shows the new DiamondMax Plus 9 SATA Hard Drive putting up some impressive figures in standard SATA 150 and SATA 150 RAID 0 configurations."
This discussion has been archived. No new comments can be posted.

Serial ATA Drives Mature and Get Faster More

Serial ATA Drives Mature and Get Faster

Comments Filter:

  • Hidden wiring/tidiness (Score:4, Interesting)

    by baryon351 ( 626717 ) on Monday March 24, 2003 @07:49AM (#5582446)
    One of the things I love about SATA is the simple clean wiring, but it's not something I see done very well in most case mods. Anyone seen images or sites related to making a truly minimal case inside? Hidden or extremely tidy floppy, CD, power and drive cables? I'd love to see how others have handled their tidy up jobs.

    • Re:Hidden wiring/tidiness (Score:2, Funny)

      by Anonymous Coward
      I'd love to see how others have handled their tidy up jobs.

      Well, depending on where you're "working" I recommend using a dry tissue to clean up the mess first and then go over the area with a damp tissue afterwards to sanitize the area. You may have repeat these steps in order to clean up the entire mess.

    • Bluetooth Wires! (or no wires) (Score:5, Interesting)

      by creamandchives ( 654511 ) on Monday March 24, 2003 @07:55AM (#5582463)
      This may sound silly, but how cool would it be to have some kind of wireless cabling system for connection between all pc devices (like bluetooth) i know its totaly inpracticle, but u could have one of those cool induction charging matts with a motherboard, hard drive, cdrom, etc just sitting on it with no wires! very trippy 8-)
      • Absolutely!. Incredibly impractical with current tech, probably highly expensive, and a bit pointless for 99% of users. But utterly cool all the same
      • This may sound silly, but how cool would it be to have some kind of wireless cabling system for connection between all pc devices (like bluetooth) i know its totaly inpracticle, but u could have one of those cool induction charging matts with a motherboard, hard drive, cdrom, etc just sitting on it with no wires! very trippy 8-)

        Until your little sister walks in and picks up the hard drive for your web server to make a good doorstop. :)

        Michael

      • While induction might be a really convenient method to power devices, you'll have to be careful how powerful your magnetic field is. With magnetic devices like hard disks and floppy disks, you'll need to be careful not to corrupt your data.

        I think it could make for really nice connectors, though. Especially in extreme environments, where dirt and grime are likely to get in between the contacts. You have something like teflon-coated springs, to serve as air-core inductors. This has the double-benefit of
      • Not impractical, just requires some standards. IEEE or whomever can say a new type of drive bay should be designed so that drives connect to power recepticles when you insert them all the way into the bays. You would then connect one set of wires from the power supply to your case's drive cage and be done. Standardize on serial ATA/SCSI, and you can eliminate data wires (or minimize them), providing hot-swappable devices.
    • Anyone seen images or sites related to making a truly minimal case inside?

      Uhhh.... Yeah. [apple.com] :-)
    • SATA is just a start. There are other things that contribute to the problem. There's the wiring harness on the power supply. That's one mess of spagetti. Also, they still need to get ATAPI supported in SATA so you can get DVD/CDROM off of parallel cables. Case design is also a factor. Apparently, wire and cable routing is not one of the issues addressed in case design. Position and layout of physical components always seems to be the worst possible from a cable layout point of view. You always seem
    • I was part owner of a clone building place once. We used to have informal contests to see who could build the neatest PC (using stock case/etc). We would clamp up custom flat cables (this was years ago, before the era of round cables), making them so that each connector was in exactly the right place, and facing the right way. We had what I thought was an innovative little feature; at the time it was almost standard to put in both 3.5 and 5.25" drives; so we put TWO controller connectors in, 1 inch apar

  • Painful (Score:5, Funny)

    by Anonymous Coward on Monday March 24, 2003 @07:50AM (#5582448)
    what models are trickling out from Seagate and Maxtor, are beginning to look promising.

    The sound you can hear is the echo of the breaking of the hearts of ten thousand grammar nazis.

  • Seems like they say wait... (Score:5, Informative)

    by 3.5 stripes ( 578410 ) on Monday March 24, 2003 @07:50AM (#5582453)
    As the tech is still pretty new, and could use some tuning. Not too surprising, most new tech seems to follow ths path.

    Does being an early adopter really have much benefit besides bragging rights?

    I was planning on waiting anyhow, this just seems to confirm my original instincts.

    • Re:Seems like they say wait... (Score:5, Interesting)

      by baryon351 ( 626717 ) on Monday March 24, 2003 @07:56AM (#5582469)
      Does being an early adopter really have much benefit besides bragging rights?

      Curiosity. Somewhat like nabbing and installing a beta copy of some software and checking out where it's at. If I had the cash and felt the satisfaction of my own curiosity was worth it, I'd have a few SATA drives running, just for the hell of it.

      It gives some geektypes something to talk about, ponder over, and throw opinions around on where stuff's heading.

      Then yes, there's bragging rights :)

      It's nothing too serious, really. People are people and some of us just like new stuff
      • Beta software I understand, you can always upgrade... and as often as not, it's free.

        Early adopting hardware seems to be a risk, as you're spending money, and you have to be pretty lucky to get it for free.

        If I had the money to burn however....
    • Nah,EMC and some others have held off for SATA 2, which will have better hot swap support and transfer rates, IIRC.

      ostiguy
    • I was planning on waiting anyhow, this just seems to confirm my original instincts. What does not kill me just postpones the inevitable.
      . People, people, people. You have to start checking for conflicts between your sig and your message. I mean, using Serial ATA might shorten the life of your data or something, but it won't kill you!
    • Someone has to do that testing. Early adopters trade the risk of the technology being crap vrs. the odds of it working. If everyone waited for the price to come down, or whatever, it'd never happen. More than one decent technology has been pushed aside because of this.

  • More Information (Score:5, Informative)

    by leibnizme ( 264472 ) on Monday March 24, 2003 @07:54AM (#5582459)
    If you want to know more about the Serial ATA technology:

    Cnet [com.com]

    SATA and ISCSI [infoworld.com]

    Intel Dev Paper [intel.com]

    Maxtor White Paper [maxtor.com]

    Serial ATA Working Group [serialata.org]

  • The size of future computers... (Score:5, Insightful)

    by march ( 215947 ) on Monday March 24, 2003 @07:56AM (#5582465) Homepage
    I once heard that the size of future computers will be limited by their component's connectors.

    That said, I wonder if we will ever get to the point of performance where a drive can sit next to a computer and communicate via a (secure) wireless connection - either RF or IR (or ??).

    Of course, then the above phrase will be that the size of future computers will be limited to their component's antennas.

    Basically, I've *got* to find some way to get rid of the huge clump of cables under my desk!

  • 'Maturing' offers more promise than 'innovation' (Score:3, Informative)

    by Anonymous Coward on Monday March 24, 2003 @08:09AM (#5582496)
    One fact I am compelled to revisit upon reading these press releases of "stronger, faster, better" technology, particularly that which is promised to be coming real soon now, is that virtually all recent advances in industry techniques have been incremental. This is not a claim that there is nothing new to be found in the business; rather, I am inclined to state that if you want to peer into the future all you need to do is apply a bit of chrome to today's offerings.

    Case in point: while stories of (distant future) storage technology consistently fill all the typical industry rags, a very real technique is already available and well-known to insiders. DVDA, one of the newer ideas that has taken off, promises to roughly quadruple conventional hard-medium storage techniques. Although more prone to tolerance faults because the scheme involves replacing the typical single-head approach with four carefully-positioned around the box, the increase in input capability has lead many to believe that consumer demand for DVDA will rise rapidly as it begins to hit the shelves in larger numbers.

    We've all chuckled over the "640K is enough for anybody" quote, but the reverse approach of industry visionaries who predict teraflops of holographic storage or similar pie-in-the-sky schemes is similarly unlikely to lead us to tomorrow's breakthroughs. Don't be fooled into thinking that we've fully exploited the potential of current techniques.

  • Slightly Off-Topic (Score:4, Interesting)

    by epicstruggle ( 311178 ) on Monday March 24, 2003 @08:09AM (#5582498)
    Now that it looks like HD manuf. are getting SATA drives out the door. Does anyone know when we could expect to see optical drives out too?

    Id love to see the end of all IDE cables in my computer. Im using a small form factor(sff) shuttle, and one of the problems with circulating air is the IDE cable. Also is there any plans/ideas about all the wires coming out of the PSU, as in any way to make those wires thinner and less obstrusive(sp?)

    thanks for any and all responses.
    later,

    • Re:Slightly Off-Topic (Score:1, Funny)

      by Anonymous Coward
      Make Use longer and softer wires, and tape/tag them down to surfaces inside the machine. If they have to wind around components on the motherboard to reach the mobo power connector, then do that.

      It works, and looks seriously cool, like some bizarre tentacled thing stretching across a vast post-apocalyptic landscape of capacitors, ICs, smt components and tracks.

    • Re:Slightly Off-Topic (Score:2, Informative)

      by uspsguy ( 541171 )
      Given the power requirements of ever faster processors, I think you may see bigger, fatter power supply cables. The physics of wire size / current capacity are pretty well established. Unless room temp superconductors become a cheap reality, plan on a wad to power the MB.
      • They could start using five 8 guage wires instead of the many small ones :)

        Eliminating the negative voltages that aren't really used anymore for much could help. Most modern implementations of RS-232 cheat and use inverted TTL anyway.

  • Why serial ATA? (Score:5, Interesting)

    by DrXym ( 126579 ) on Monday March 24, 2003 @08:12AM (#5582503)
    Out of curiosity, what is the point? Firewire presumably offers comparable performance doesn't it, or is there some compelling reason not to use it such as lower bandwidth or contention issues? And firewire seems to be a standard feature on an increasing number of PCs these days.


    Now, I would welcome any replacement to conventional IDE / ATA which has been the bane of my life. I couldn't count the number of times I've had to screw around swapping cards and drives in order to accomodate that ribbon. I will be happy to see that particular technology go the way of the dodo.

    • Re:Why serial ATA? (Score:5, Insightful)

      by Psiren ( 6145 ) on Monday March 24, 2003 @08:35AM (#5582561)
      I think the main thing (feel free to correct me anyone) is that SATA is a drop in replacement for Parallel ATA. It should just work, without requiring a whole bunch of new drivers. That's not to say that writing a whole bunch of new drivers wouldn't get more out of the technology, but it's not a requirement in order for it to work. Firewire on the other hand does require special drivers, and it's not yet *that* common on desktop PC's. Also, although I'm not sure of the state of Firewire support in Linux, I would bet anything it's nowhere near as good as the IDE support (which still has its own problems of course).

      • Re:Why serial ATA? (Score:5, Interesting)

        by DrXym ( 126579 ) on Monday March 24, 2003 @10:24AM (#5582935)
        I think other responses explained the general reasons that serial ATA might be better than firewire but I thought I'd make one point. The absence of a driver in Linux is not a reason in itself to not adopt a technology. Linux gets a driver when there is a demand and motivation for such a driver. The emergence of snazzy firewire drives, camcorders etc. is exactly such a reason and will fuel development of such features. The same happened when USB first appeared.


        As far as I'm aware IEEE 1394 (firewire) is a readily available standard and assuming specific chipsets are documented there should be no barriers to making Linux talk happily with such devices. I'm no kernel engineer but I would guess that great big chunks of their bus / device abstraction are readily applicable to firewire too.

    • Re:Why serial ATA? (Score:5, Informative)

      by NomNet ( 542670 ) on Monday March 24, 2003 @08:51AM (#5582595) Homepage
      Out of curiosity, what is the point? Firewire presumably offers comparable performance doesn't it, or is there some compelling reason not to use it such as lower bandwidth or contention issues? And firewire seems to be a standard feature on an increasing number of PCs these days.

      Firewire gives a maximum throughput of 400Mb/second (50MB/second), with future versions giving 800Mb/second (100MB/second).
      SeialATA gives a maximum throughput of 150MB/second, with future versions giving 300MB/second and then 600MB/second.
      SerialATA is MUCH faster. Now granted, modern Hard Drives can't get anywhere near 150MB/second, but one day they will :)

      • Re:Why serial ATA? (Score:2, Interesting)

        by ed1park ( 100777 )
        This reminds me...

        http://www.tomshardware.com/mobile/20020827/inde x. html

        A review of USB 2.0 and Firewire drives at Tom's site indicates that real world throughput is bottlenecked by the interface to about HALF the theoretical max.

        Why so slow? Is it all overhead? Poor optimization? Should we always assume that real world performance is approx half that of the theoretical max?
    • Firewire is designed as an external interface, and as such is limited by long cable speed etc. Serial ATA is an internal only interface, and can achieve much faster speeds. It's also a point to point system, so it doesn't suffer from Firewire's need for device identification etc.

      Also, Firewire is designed to supply power over the interface - Serial ATA dedicates lines for this, which aren't optional.

    • Re:Why serial ATA? (Score:5, Informative)

      by samael ( 12612 ) <Andrew@Ducker.org.uk> on Monday March 24, 2003 @08:55AM (#5582608) Homepage
      Firewire can be connected in a multitude of different ways, to different devices. It therefore needs a fairly complex protocol.

      ATA can be connected in very few ways to only one controller. It therefore has a nice, simple protocol.

      The simpler the protocol, the higher the throughput, because you're not having to send messages and wait for replies to work out where things are going.
    • Firewire1 maxes at 40MB/s SATA1 is 150MB/s, Firewire2 is 80MB/s SATA2 will be 300MB/s. Also the electronics for SATA are relativly simple, in fact once they are up to volume they will probably costs LESS than current PATA chips whereas Firewire is kind of expensive. Firewire is good for external drive enclosures and as a transport for timing senesitive data (such as video in a pc-less setup or next generation MIDI which will use Firewire for both note and sampling data)

  • solid state? (Score:3, Interesting)

    by khuber ( 5664 ) on Monday March 24, 2003 @08:18AM (#5582516)
    I'm still waiting for affordable solid state disks. Magnetic storage seems so crude.

    -Kevin

    • Yeah! Damn credit cards!

      • Re:solid state? (Score:2, Insightful)

        by khuber ( 5664 )
        Yeah! Damn credit cards!

        Exactly - credit cards have been using the same magnetic strip "technology" since the early 70s.

        Japan and Europe have smart cards in wide use, but the U.S. lags. We still manually input name and address information into disparate point of sales systems for every vendor. That's just sad.

        -Kevin

        • We still manually input name and address information into disparate point of sales systems for every vendor.

          Uhhh... We still input our pin numbers as well. I wouldn't consider that a bad thing. Inputting an address that is required to match with the CC company's record is the best basic security, and the only security that CC have right now.

          You've got the same mindset as all the "experts" that say that Japanese consumers are so far ahead of the USA in technology... as if buying another crappy little use

    • How quaint. (Score:2, Funny)

      by fbg111 ( 529550 )
      I'm waiting for quantum optical storage cubes. Solid state seems so crude. ;)
    • crude is hand feeding magneticaly striped ledger cards into a NCR 500 for mass storage, and a deck of punch cards for inputing the transactions.

  • What about drive failures? (Score:5, Interesting)

    by jspectre ( 102549 ) on Monday March 24, 2003 @08:57AM (#5582615) Journal
    Ok. It's nice to see new technologies getting out there for hooking drives up and making them lickety-split fast. But in the past year or two I've purchased 20something hard drives of various sizes from leading manufacturers and had AT LEAST one drive from each fail, if not two or more. This includes Quantum, IBM (who smartly got out of the business), Segate, Toshiba and others.

    How about someone making a hard drive that isn't going to give up after a year? Or are these guys only in the business to sell me new hard drives after a year? Many are also reducing their warranties from 3 or 5 years to one year. Have they no faith in their own products?

    • Re:What about drive failures? (Score:4, Interesting)

      by nmg196 ( 184961 ) on Monday March 24, 2003 @09:24AM (#5582706)
      How about someone making a hard drive that isn't going to give up after a year?

      I think I have a solution for you - or rather Western Digital do. It's 10,000 RPM, 5.2ms average seek time, SerialATA 150, 1.2 million hours MTBF, and a 5 year warranty! With those stats, it should really fly if they've put some effort into the controller...

      It hasn't been released yet, but I'm going to wait and see what the reviews say before upgrading my machine - it looks good on paper.

      They have a press release here [westerndigital.com].

      Nick...

      • Re:What about drive failures? (Score:1)

        by Anonymous Coward
        COOLING. I believe is you see these failures cooling might be something you should look into.
      • Maxtor has similar plans with their MaxLine II P + SATA Drives. Not quite as impressive as WD promises, 7200 Max RPM, 3 year warantee, but a claimed MTBF that is as high as their SCSI line.

        MaxLine II should be less vapor than the WD offering at least, they had planned to have it out 1Q 2003, but it seems they are still hard to find. Some places are accepting orders, however.
      • 1.2 Million hours? Sure, I know some drives do last a long time but 137 years? And I certainly wouldn't put my money on a fast, hot, 10,000RPM drive lasting that long either.

        I would like too see any relatively complex machine last 137 years without repair, even under ideal conditions. Especially something as sensitive as a hard disk.

    • Re:What about drive failures? (Score:4, Interesting)

      by Surak ( 18578 ) <surakNO@SPAMmailblocks.com> on Monday March 24, 2003 @10:04AM (#5582843) Homepage Journal
      From the article:


      Although clearly this is a high end product, the warranty from Maxtor for this type of drive is lack-luster. Maxtor only warranties the drive for 1 year, while many other high end desktop products, like the WD SE drive, sport 3 year warranties. We would like to urge Maxtor to reconsider upping the anti here. Clearly this is a well built product, there is no reason not to stand behind it for a longer duration
      /blockquote

      The reviewers clearly felt that the Maxtor SATA drive was well-built and that there was no reason that Maxtor SHOULDN'T stand behind it for longer.

      Personally, I think that hard drive manufacturers aren't standing behind their drives for longer for a few reasons. One is that margins are razor thin. You'll be able to pick up an 80GB Maxtor SATA-150 drive for $80-$125, once they begin to ship in quantity. Just a few years ago, the top-of-the-line ATA drives (whatever their size is is irrelevant..I'm talking line positioning here) were selling more than 3 times that price. Inflation has gone up, but technology has deflated in price quickly. (Personally, I'm waiting for them to come free in my box of Lucky Charms(TM). :-P) So if they have to warranty their drives for 3 years, they may not be able to stay in business.

    • Get SCSI (Score:3, Interesting)

      by spanky1 ( 635767 )
      No, I'm not trying to start another flamewar. IDE vs SCSI is an old argument. But one thing is certain: SCSI drives are much, much more reliable. They are designed for server use (24x7x365) and can be much faster. Get the Cheetah 15k.3 and you'll never look back!

      Notice: You will pay for SCSI reliability.

      • Re:Get SCSI (Score:5, Funny)

        by nmg196 ( 184961 ) on Monday March 24, 2003 @12:12PM (#5583521)
        > Get the Cheetah 15k.3 and you'll never look back!

        Warning: Never put Seagate Cheetah drives where there are people. They sound like a circular saw trying to cut though a piece of reinforced concrete. Really not a very nice sound. At all.

        Nick...
        • Apparently you have never heard a Cheetah 15k.3. It is VERY quiet... barely audible. Also, it runs extremely cool to the touch. Your observation was accurate for first gen 15k drives, but not current 3rd generation ones.
          • No I confess - it was an old Seagate Cheetah 10K. It was most unpleasant.

            If you want quiet/fast drives anyone (normal ATA) I highly recommend the Western Digital JB range (8mb cache versions). You can hardly hear them and they seem to be very quick compared to my IBM equivalent. It might be psychological though - now that I don't get a grinding noise everytime I do something, everything seems much quicker :)

            Nick...
            • Yeah, those WD JB drives are very quiet. I have a 120GB drive in my system for "junk" like MP3s and games. My Cheetah 15k.3 is a hell of a lot faster though so I use it for my system/app drive. The WD JB was cheaper and has 4 times the capacity though. :) But I expect the Cheetah to outlive it by quite a bit.

  • Toms Hardware (Score:5, Informative)

    by nmg196 ( 184961 ) on Monday March 24, 2003 @09:04AM (#5582639)
    Toms Hardware also has posted a review [tomshardware.com] of this Serial ATA drive.

    Summary: "Extremely High Performance, Excessively Short Warranty Period"

    Nick...

  • Already? (Score:2, Interesting)

    by imaniack ( 638051 )
    I bet 99% of computers does not even have one single SATA drive and it's MATURE already? Where are the beta testers, ummm I mean cutting edge users?
  • Why is it called serial? As far as I can tell, only one drive can be connected per port on the motherboard. Why not "single" ATA? When it's called serial, I envision something like SCSI, with the ability to daisy-chain multiple drives.

    • Re:Just one question... (Score:5, Informative)

      by SuiteSisterMary ( 123932 ) <slebrunNO@SPAMgmail.com> on Monday March 24, 2003 @11:31AM (#5583301) Journal

      In the days of yore, when you could send only a few bits per second down a wire, which is serial, it was noticed that you could lump eight wires side by side, send one part of a byte down each wire, and boom, you've got parallel. Like this:

      Serial:
      0
      1
      1
      0
      0
      0
      1
      0
      Parallel:
      001100010

      Now, however, they've noticed that our ability to send bits down a wire is so improved, you're actually wasting time by trying to synchronize between eight separate wires; it's faster to just blitz the 8 bits down one wire.

      Hence, this new ATA is serial, whereas (E)IDE is parallel (those flat ribbon cables give it away nicely, don't they?)

  • About ATA drives (Score:4, Informative)

    by mr_stark ( 242856 ) <[tim] [at] [trgray.co.uk]> on Monday March 24, 2003 @12:34PM (#5583700)
    What Does ATA stand for?

    ATA = AT Attachment

    AT = Advanced Technology (as in IBM's first PC)

    Basically the old IDE (Integrated Drive Electronics) and the later UDMA (Ultra Direct Memory Access) drives are parallel ATA devices, that is data is sent over multiple 'lines' Serial ATA send data over 1 'line' but at a much faster rate.

    In theory parallel transmission should be faster, more lines = more bandwidth but in practice serial connections are quicker as they don't suffer from cross talk and other complications (big cables - easily damaged)
    .
  • During a PCWORLD article I read up on the difference between SATA and the PATA. While Serial was able to do better on searching and moving a bunch of smaller files. Parallel was still able to beat Serial in opening a single 1.7 gig file. And the difference in some cases was only by a few seconds. This, to me, doesn't want to make me change anytime soon. I don't see this coming into play for a few more years when they are able to make the transfer's faster for big files or they just decide to change it
  • Given the mobo manufacturers' tendency to leave legacy connectors on long after the need for them has all but disappeared, i.e. parallel & serial connectors, I suppose we'll still see PATA connectors on the mobo's for years.

Slashdot Top Deals

"When the going gets tough, the tough get empirical." -- Jon Carroll

Close