Slashdot Mirror


Next Wave Of Hard Drive Tech: Perpendicular Recording

angrytuna writes "New serial technologies are set to replace standard SCSI and ATA (Advanced Technology Attachment) interfaces over the next two years, even as hard-disk drive manufacturers prepare for an entirely new form of bit storage. Perpendicular recording will replace longitudinal recording in storage devices, placing bits on end instead of lying them parallel on the disc surface, thus dramatically increasing the possible storage density."

25 of 380 comments (clear)

  1. Comparing apples and hydrogen by Kris_J · · Score: 2, Insightful

    Why does the article reference interfaces then talk about a new way of storing the bits on the disk survace?

    1. Re:Comparing apples and hydrogen by AllenChristopher · · Score: 4, Insightful
      It talks about new interfaces because we've been hearing so much about interfaces that it's a familiar topic to which readers will relate, and which can form a bridge into the story. It's a variety of lead paragraph.

      An article which simply jumps into a description of an esoteric subject can seem awkward and be difficult to understand, so journalists have long been taught strategies for lessening that initial impact. Many of these conventions don't play as well in the internet environment because a linking page has already told the reader what the article will really be about. This makes the lead seem like irrelevant wandering.

  2. Details? by Christopher+Thomas · · Score: 4, Insightful

    Does anyone have a link to a description of this that's more detailed than "stacking bits on end"?

    Are they using platters with trenches and storing information on the sidewalls?

    Are they using some means of reading and writing at many depths within the platter without disturbing other layers?

    The article says the technology has been under investigation for 20 years, so presumably there's a forest of technical literature on it.

  3. Density doubling annually; access speeds lag by Allen+Varney · · Score: 5, Insightful

    This conversation with Jim Gray, head of Microsoft's Bay Area Research Center, has grim, eye-opening comments on the growing gap between storage densities and access speeds/bandwidth. Currently the most effective way to send a multi-terabyte disk array is by UPS -- turns out a UPS truck has a "bandwidth" equivalent to about 7 megabytes/second. And the problem of practical access speeds is only going to get worse. At current and near-future access speeds, searching a 20-terabyte disk might take a year.

    "At the FAST [File and Storage Technologies] conference about a year-and-a-half ago, Mark Kryder of Seagate Research was very apologetic. He said the end is near; we only have a factor of 100 left in density--then the Seagate guys are out of ideas. So this 200-gig disk that you're holding will soon be 20 terabytes, and then the disk guys are out of ideas. The database guys are already out of ideas!"

    1. Re:Density doubling annually; access speeds lag by pjrc · · Score: 2, Insightful
      Quoting the parent post, who was quoting Jim Gray...

      At current and near-future access speeds, searching a 20-terabyte disk might take a year.

      Today's drives run about 20 to 50 Mbyte/sec from the platter. You can get 133 Mbyte/sec from the tiny buffer, of course, but for a whole-drive search, let's assume you're going to read 20 terabytes at 30 megabytes/sec. My calculator says that's 666667 seconds, or 7.7 days. Yes, a long time to wait, but 7.7 days is a long way from "might take a year". Even if you get only 10 Mbyte/sec, which is much slower than the drive's benchmark'd transfer speed but might happen with operating system and other other overhead, you're still at one month, not one year. To take an entire year reading 20 TB, you'd need to sustain only 634 kbyte/sec transfer. Hard drives haven't been that slow for a very, very long time.

      I'll be Jim Gray (head of Microsoft's Bay Area Research Center) believes what we all need is to dump FAT32, NTFS (ext2, ext3, JFS, Reiser, XFS, etc) and go to whatever database oriented filesystem Microsoft is cooking up for 2005 in "longhorn"... or whatever they're naming it nowadays. The sad part is that it'll require faster CPU and hard drive to achieve the same performance of older MS system, assuming they continue the long running trend of "innovation".

  4. Density is nice, but I need speed! by smokeslikeapoet · · Score: 5, Insightful

    Storage density is one thing, but storage speed is another. With 200 GB hard drives readily available, and relatively cheap, the main thing I'm itching for is increased access and transfer speeds. Not just the controller speed as most hard drives still only maintain a constant transfer speed of 33Mbps. Theoretically, a denser drive at the same rotational speed will transfer data faster than a less dense drive, but will we see a dramatic improvement in sustained transfer speeds? While this transfer speed is acceptable while watching a DivX movie, it's really a pain while ripping a DivX movie. (A movie that I shot in my backyard, and authored, and own the rights to, and am ripping for the pure exitement as I would never violate a copyright.)

  5. Re:a shame then by madsenj37 · · Score: 3, Insightful

    Not having to compress video and audio, thus not degrading the quality, is one use we would not mind having. It is good for both pros and average users alike.

    --
    Choosing the lesser of two evils is a choice for evil.
  6. Re:Increased Reliability? by l810c · · Score: 5, Insightful
    Not to mention backing up that data. Why does backup technology lag so far behind drive technology? 100+ CDR's to backup the typical Hard Drive in today's systems. 25 DVD's, but still WAY too much. A tape sytems to backup the typical hard dirve in a reasonable amount of time costs in the thousands.

    I'd like to see Redundancy And Speed hit the consumer market more than the current volume. RAID 0+1 should be standard in at least mid level systems.

  7. Re:And? by SpryGuy · · Score: 4, Insightful

    Well, higher density means the same storage in a smaller form-factor, which means the read/write heads have to travel a smaller distance (both radially and logitudinally), which should yield a measurable boost in potential performance... no?

    --

    - Spryguy
    There are three kinds of people in this world: those that can count and those that can't
  8. das shrunken by poptones · · Score: 4, Insightful

    More relevant than this technology that is still many years away, I find much more interesting the part about the desktop industry moving to 2.5" drives. So in the next year or so we'll be able to buy very high density, fast drives that can fit in a pocket and already have serial interfaces! All we need are sata jacks on the front panel and the world moves one giant leap closer to true "plug-n-play" goodness. Mail order sneakernets just got even cheaper!

  9. Re:a shame then by James_G · · Score: 5, Insightful

    I just stopped on the way home and did some photo shooting. I took 57 photos in about an hour. At 7.2MB per shot, that amounts to ~414MB of files from just an hour of shooting.

    Post-editting results in TIFF files that are approximately 10MB in size. All told, this one shoot now occupies over 800MB on my fileserver - from just one hour of shooting.

    Oddly enough, people do in fact use vast amounts of storage space for reasons other than sharing mp3s and movies. As technologies improve (cameras increase resolution, video cameras likewise, millions of other reasons), the demand for space will increase as it always has done.

  10. Re:hmmm by UniverseIsADoughnut · · Score: 2, Insightful

    I'm curious to, but i think it's something like this. If I have a log laying on the ground i can rotate it around to put it in it's proper place. But if i stand it on end and rotate it about its axis to get it in its differant positions its not covering nearly as much space. But keaping it standing up is the problem since it naturaly wants to fall over, and you can't attach the base of it to the ground very well if your going to be spinning it all the time.

    This is just my thought on what there doing. But how data is actualy stored on a disk is a mystery to me, all i have is guesses. Soon as i think that one little thing not going the right way destroys a file causes me to think it must be weird magic voodoo.

  11. Back in the day by appleLaserWriter · · Score: 4, Insightful

    One of the first Sun machines I used was a 3/160 with an external gigabyte disk array. The array was a washing machine size enclosure with a pair of 800 MB SMD disks with 8" platters. In 1994 this was a huge disk, in more ways than one!

    Interestingly, my little 486 with its 340 MB drive were far faster than the old Sun, and even competitive with the newer SparcStations. 7200 RPM baraccudas in modified enclosures (extra fans and breathing holes made the difference between life and death) were even faster when they arrived.

    After working exclucively with laptops for the past two years, I can see a clear parallel between the old 2.5" -> 2.5" transition and the 8" -> 5.25" -> 3.5" transitions in the past. Sure I keep a pair of 120 GB 3.5" disks in firewire enclosures around, but the 60 GB disk in my powerbook and the 30 GB disk in my Dell i8000 are more than adequate for daily use. My ipod even has 30GB, which is enough for my favorite music, the Warthog Jump video and a few other fun things.

    With emphasis on blade and 1-U servers, as well as cardcage oriented telecom gear, I can see a lot of value for 2.5" disks in the telecom and server markets.

  12. Re:Increased Reliability? by evilviper · · Score: 4, Insightful
    Why does backup technology lag so far behind drive technology?

    You know, drive technology IS backup technology. Just backup data to hard drives instead of tape or CDs. Also, I believe a big reason is the density of data. No longer can you have something like floppies, where punching a tiny hole in the media won't cause a problem. Now, a speck of dust making a tiny microscopic scratch would ruin megabytes of data on a HDD platter, so they can't make unsealed media like DVDs or tapes that dense, now can they?

    I personally never understood why sealed media never got popular. CDs with caddies would be far better, but people stuck with bare-assed, easily damaged CDs instead. Same problem with DVDs. Minidiscs aren't very popular unfortuanately.

    It's possibly that tightly sealed media could be much higher capacity than currently seen, but who's going to be the one who suggests to their boss that they should try doing something that has failed every other time it has been tried?

    Zip-style disks could potentially provide very high capacities, but they can't expand as quickly as hard drives... To do that, you'd need someing with it's own controller, like CompactFlash or hard drives.

    RAID 0+1 should be standard in at least mid level systems.

    I've said it before, and I'll say it again: "RAID is not a backup technology." When your main disk gets hosed by a virus, a clumsy user, or a system crash, that corruption is coppied to the other disc at light speeds... So what's the point? Offline backups are what is needed. RAID provides a solution for hardware problems, which is important with critical systems, but if the hard drive in my home PC crashes after a year, as long as I can restore a recent backup, and only be down for a few hours, it's not really a problem.
    --
    Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
  13. Re:Increased Reliability? by Dragoon412 · · Score: 2, Insightful

    I couldn't agree more...

    Let's play a game of Find the Bottleneck!

    Pentium 4 3.0C 800mhz CPU: 6.4 GB/s bandwidth
    Sapphire Radeon 9800 Pro GPU:: 21.8 GB/s bandwidth
    PC3200 DDR-400 RAM: 3.2 GB/s bandwidth
    ATA/133 Hard Drive: 133 MB/s bandwidth

    Now, I'm an avid gamer, and I fully realize that, comparing the performance, of, say, a video card to a hard drive isn't fair, because that video card has to do a lot more work than the hard drive, in a gaming environment. Still, when the video card's capable of pushing in the ballpark of 200 times the data that the hard drive is, it's quite telling of a deficiency.

    We already have more storage than virtually anyone knows what to do with. How about making the drives faster?

  14. Sag by leonbrooks · · Score: 2, Insightful
    If you store 8 bits deep, you can read a byte in the same amount of time it takes to read a bit.

    AFAICT, they're not talking about multi-layer recording, they're just standing the existing bits on end so that the same amount of magnetic material uses up less surface real-estate. <deadpan>If they did multi-layer recording, they'd have to slow the drives down so that the surface of the disk wasn't so stretched by centrifugal "force" and the shallower bits didn't sag into the next cylinder. Otherwise they'd have to angle the heads WRT the platter surface, which means they can't fly them close enough to record that deep.</deadpan>

    --
    Got time? Spend some of it coding or testing
  15. 2.88M by atcurtis · · Score: 2, Insightful

    Weren't the 2.88MB floppy discs perpendicular recording also?

    I distinctly remember reading that somewhere.

    --
    -- The universe began. Life started on a billion worlds...
    -- Except on one where stupidity was there first.
  16. Re:Windows Longhorn by SN74S181 · · Score: 2, Insightful

    People seem to forget that if the media industries (all hated virulently on slashdot, it seems) are going to make money in the future with digital transfer, they'll need a means to shove content onto our local drives so we can watch it from there. In fact, one of the things that makes me suspicious regarding this new quantitative leap in storage... is that these drives might be DRM enabled in hardware.

    Yes, even if P2P is banned somehow, these high capacity drives will be needed. From the point of view of Hollywood, it is imparative that they be widespread, so they can shovel content onto them and charge us their appropriate fees for doing so.

    Not saying it's good or bad, just that it is.

  17. Re:Can we get real here by brarrr · · Score: 2, Insightful

    well, this is one article about one idea. there are many more facets of magnetism research. i'm working on diluted magnetic semiconductors that put ~1ppm magnetic dopant into a semiconductor... this can conceivably allow for solid state, non-volatile, always on, (insert your term), memory. perp. recording has been around for a while, just not economically feasible, and who knows when it will be.

    there is no company who is reaching 1terabit per cm2, as there is something called the superparamagnetic limit

    sorry to bitch and moan, but it bothered me..

    --
    to email me: take my /. handle and append .net preceded by charter.
  18. Re:Increased Reliability? by Fweeky · · Score: 2, Insightful
    as harddrives get more and more high-tech, the reliability seems to be taking a big nosedive

    Not really. HD's are now more reliable than they've ever been; they're just a lot more common (9 HD's in my various systems at home now), and often not treated with the respect such precise electronics need.

    Tips to making sure your HD's at least reach their design life:
    1. Don't buy from that cheapo supplier who's boxes are always a bit mangled. HD's don't like shock, even when they're off.
    2. Keep your drives cool. Especially now 7200RPM is so common, drives frequently report temperatures rivaling that of CPU's. Heat is not good for drives or their electronics, and a lot of cases have their 3.5" bays in air deadzones. A slow, effectively silent fan could mean the difference between a 3 year lifespan and a 7 year one.
    3. HD's are fragile. Treat them like eggshells; don't force them into that tight little bay on your cheapo case and end up ramming it into the front of the bay. Duh!

    Common sense, really...
  19. What a poorly researched article by pbox · · Score: 2, Insightful

    What a poorly researched article. It is way below EEtimes quality and should have never been published as it is.

    1. "As opposed to longitudinal recording, where the bits are impressed in a parallel format along the surface of a disc, perpendicular recording stands the bits on end, enabling more data storage per square inch."

    What does it stand the bits at the end? I have never seend a standing bit. Especially on the end of it. Now c'mon, it could have been described a little more "technically". This is not USA Today.

    And "impressed in parellel format" is such a crap of a phrase. It is not impressed, nothing touches, no impression, it is MAGNETIC, god damn!

    2. "Apple Computer Inc.'s new G5 computers are all SATA-based while Intel systems will by the end of this year be based on the new interface."

    Now this is utter turd. Before even G5 was announced, and probably before Jobs had the brainfart to invent them, some of the high-end PC motherboard manufacturers were already churning out SATA equipped motherboards. It was in the Intel development road map for several years now. I remember reading about it on Tom's 2 years ago.

    Mr. Bolaji Ojo (EBN), please do your homework. Do not just blair (as in Jayson) out an article. You do wipe your ass after takin a sh*t, don't you? I am just asking that you would apply the same attitude toward writing articles. Thank you for your future cooperation.

    --
    Code poet, espresso fiend, starter upper.
  20. Re:hmmm by panurge · · Score: 2, Insightful
    Magnetic fields interfere with one another. If the tiny magnets that represent 1s and 0s get too close they can depolarise one another (=loss of signal).

    Think of matchsticks flat on the floor and standing up. The ones standing up will be further apart, or you could pack more in the same floor area and have them the same distance apart.

    The difficulty is, matchsticks have an easily distinguishable top and bottom end, but magnet ends are only distinguishable by the direction of flux. The bottom ends are buried in the media, so there are challenges in writing and reading the data because only one end of the magnet is accessible. Does this help?

    --
    Panurge has posted for the last time. Thanks for the positive moderations.
  21. Re:Increased Reliability? by Mr.+Sketch · · Score: 2, Insightful

    Hard drives are _not_ a backup technology. The whole point of tapes/cds/dvds/etc is to decouple the data storage from the reader. Thus the data is often stored in a solid state medium making it less succeptable to failure while the reader often contains lots of moving parts making it more prone to fail. So, if the two are decoupled, when the reader breaks, I just get a new one and my data is safe. But if we 'backup' on a hard drive, and the hard drive crashes or the internal hard drive reader crashes, we can't simply get a new reader, the data is just gone.

    A RAID array could be considered a backup tech if the array was treated as a backup device like a cd/dvd/tape and not just a RAID of the main 'live' system. You could still run into the problem of a drive in the array failing, but since the data is replicated, you're safe, just replace the drive, re-sync and continue on your way.

  22. Re:Increased Reliability? by evilviper · · Score: 4, Insightful
    The whole point of tapes/cds/dvds/etc is to decouple the data storage from the reader.

    No, it is just a techical decision to do things that way... Hard drives are too bulky and expensive for small ammounts of storage, and tape drives are too expensive to have one tape per drive, not to mention the bulk.

    Thus the data is often stored in a solid state medium making it less succeptable to failure while the reader often contains lots of moving parts making it more prone to fail.

    In the real world, that isn't really the case. Your tapes are more likely to be damaged than a hard drive, mainly because the hard drive is extensively sheilded. Moving parts are only a problem after a long long run-time. If a hard drive's mechanical parts were working when you stored it, it will almost certainly work when you need to recover from it... Not to mention that hard drives CAN have everything but the platters swapped if you can find an identical device, or can be recovered manually by any simple data recovery center.

    A RAID array could be considered a backup tech if the array was treated as a backup device

    If you're talking about a live backup, you shouldn't be. One power surge could take out an entire RAID array. If you are talking about off-line, I have no idea why you bring up RAID.

    Backup tapes are known to fail as well... That's why you make two of each, and send one off-site. Even if you aren't that stringent, your backup scheme should certainly have a LOT of redundancy in it, no matter what your media. I would certainly bet that hard drive failures are far more rare than tape failures.
    --
    Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
  23. Re:how bits on end would work........ by sexylicious · · Score: 2, Insightful

    Yes, perpendicular to the plane of the platters. And just a technicality: there are no such things as magnetic particles. If you think so, cut that magnet in half and you'll always get another North-South pair. ;)