Slashdot Mirror


Flash Memory to Rival Hard Drives

Skal Tura writes "Samsung will start producing 16 gigabit Nand Flash chips this year, nudging the memory technology towards use in notebook PCs and maybe even edging out hard drives in some products in the next few years."

407 comments

  1. Gb or GB? by TripMaster+Monkey · · Score: 5, Informative

    Some more information about the NAND flash memory can be found here.

    One nice thing about this article is that it clearly explains the difference between a gigabit (Gb) and a gigabyte (GB)...something the article referenced in the story seems confused about.

    From the article referenced in the story:
    Samsung will start producing 16 gigabit Nand Flash chips this year...

    Gartner estimates that 16GB Flash drives will cost from about $90...

    And from the article referenced above:
    Memory chips are measured in gigabits, or Gb, but consumer electronics manufacturers talk about how many gigabytes, or GB, are in their products. Eight gigabits make a gigabyte, so one 8Gb chip is the equivalent of 1GB.

    Sorry to be picky, but I'm a stickler for detail.
    --
    ____

    ~ |rip/\/\aster /\/\onkey

    1. Re:Gb or GB? by Anonymous Coward · · Score: 3, Informative

      Also some good insight on NAND here.

    2. Re:Gb or GB? by product+byproduct · · Score: 2, Insightful

      The difference between gigabit and gigabyte needs to be explained on Slashdot about as much as the difference between the Moon and the Sun needs to be explained to astronomers.

    3. Re:Gb or GB? by pdbogen · · Score: 4, Interesting

      The missing element here is that most flash drives, especially something in a hard-drive form factor, will have more than one flash chip. The news here is the new (much?) higher density flash chips.

    4. Re:Gb or GB? by Jozer99 · · Score: 4, Insightful

      The problem as I see it is not really chip density, but cost. If you think of the size of each of these chips, you could easily fit 60 or so GB into a 2.5" drive shaped device, and 100's of gigs into a device the size of a 3.5" drive. The problem is that these devices would cost astronomical ammounts. If we could make 1GB flash chips that cost $5, then you could have $300 30GB flash drives.

    5. Re:Gb or GB? by SavvyPlayer · · Score: 1

      This is not the first Gartner article to pass idle, uninformed speculation for hard research these past few weeks:

      http://www.gartner.com/DisplayDocument?doc_cd=1373 23

      Are there morale issues at Gartner? Drug problems?

    6. Re:Gb or GB? by Skal+Tura · · Score: 1

      I'm really sorry for forgetting to interpret this! I was quite tired and sleepy when i posted this one..
      Oh and never forget the "Marketing Factor", that 1Kilo is 1000 instead of 1024. I don't know does this affect Flash drives but i bet they do.

      But damn it's nice to see your posting on slashdot first thing at the morning :)

    7. Re:Gb or GB? by Feyr · · Score: 4, Insightful

      in some applications this is actually not a factor.

      i could even see myself replacing my OS disk with a flash based one, and have a secondary larger hard drive for the less-accessed files with gobs of ram. that would be a real blessing to my poor ears! give me a 4gb flash drive and i'll be all over it!

    8. Re:Gb or GB? by Jozer99 · · Score: 4, Interesting

      You would be supprised. Although linux can be built to use a flash (ie minimum writes to firmware) drive, Windows cant, without using Windows CE. It would be nice if Vista supported such ideas. The problem with windows is that many programs install system files. I installed XP on a 9.1 GB scsi disk, with an 80 GB IDE disk for everything but the OS. Even though I installed all programs on the 80GB disk, the 9.1GB disk was full within a year, as MS Office, Photoshop, and other stick stuff into your windows install.

    9. Re:Gb or GB? by cdrudge · · Score: 2, Insightful

      I beleive you are confusing flash memory with some form of read-only memory (EEPROM, EPROM, etc). Windows can be installed to a flash memory device, such as a CF memory card via a IDE adapter or a solid state hard disk for instance. It use to be that it wasn't recommended to use the flash card as your swap space as the constance reading/writing would eventually "wear out" the card, but apparently it's not as much of an issue as it use to be.

    10. Re:Gb or GB? by Anonymous Coward · · Score: 0

      i don't know about that - judging by some of the comments i've see on /. there are some very non-tech-literate readers here....

    11. Re:Gb or GB? by Anonymous Coward · · Score: 0

      You just ruined a moment! Shame on you!

    12. Re:Gb or GB? by Z34107 · · Score: 3, Interesting

      Bah. Stupid n00bs. I was in awe when my 80486 machine could, at long last and at great expense, support a whopping 550 MEGAbytes of FAT16 bliss! It was the size of a brick, and pretty dense, too, if I'm not mistaken. Of course now, I carry around more in a device so small that it's not a mere choking hazard, but an inhilation concern should anyone inhale too deeply around it.

      As for cost, right now they're being used in conjunction with existing hard drives as extra large buffers, so that anything "written" to the HDD very rarely needs to cause it to spin up.

      --
      DATABASE WOW WOW
    13. Re:Gb or GB? by 6*7 · · Score: 3, Interesting

      My guess is that the parent is trying to point out that Linux has filesystems like JFFS2, which try to prevent wear of sectors: http://sourceware.org/jffs2/jffs2-html/

    14. Re:Gb or GB? by 6*7 · · Score: 1

      The "marketing factor" you mention is BS, harddisk manifacturers are just about the only ones out there who use the correct SI prefixes for storage devices.

    15. Re:Gb or GB? by phookz · · Score: 0, Offtopic

      Wait a minute - I'm no astronomer, but there's a difference between the moon and the sun?!?!

    16. Re:Gb or GB? by Anonymous Coward · · Score: 0

      i think you give /. readers a bit too much credit.

    17. Re:Gb or GB? by penguinoid · · Score: 5, Funny

      The difference between gigabit and gigabyte needs to be explained on Slashdot about as much as the difference between the Moon and the Sun needs to be explained to astronomers.

      Judging by that +5, insightful, I'm tempted to make a snide remark about the ruling class (moderators). Why yes, I do have karma to burn.

      --
      Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
    18. Re:Gb or GB? by dnoyeb · · Score: 4, Informative

      Of course there is also that 'number of writes' issue.

    19. Re:Gb or GB? by Mr.+Underbridge · · Score: 1

      Once, you would have been correct, but I think that day is past.

    20. Re:Gb or GB? by Anonymous Coward · · Score: 0

      Duh. The sun is out in the daytime, the moon is out at night.

    21. Re:Gb or GB? by Firehed · · Score: 5, Interesting

      Doesn't stop the article from using the wrong one, though. I always question something that seems a bit out there. So they will indeed be 16 gigabyte flash drives? My uncle keeps going on and on about his "80 Gigabit" hard drive. Boy could I nail him on ebay reselling 4Tb hard drives. And eight gigabits of ram on one module. And... oh, why not... the 11.5 megabit floppy disk.

      --
      How are sites slashdotted when nobody reads TFAs?
    22. Re:Gb or GB? by Skal+Tura · · Score: 2, Interesting

      SI Prefixes?

      Now, what I AM talking about is that the computers world use 1,024 as 1KILO, not 1,000. Now that's the marketing factor, inexperienced people expect to see 80Gb hdd, what do they get? Roughly 74,5Gb.

      Now the rest of the computer world usees 1,024. Now tell me marketing has nothing to do with that?

    23. Re:Gb or GB? by Firehed · · Score: 1
      Well, seeing as that same $300 will buy you the four gigs of ram that Gigabyte's i-Ram can hold (not even the card itself), it's actually not too bad of a price. Except the whole read/write lifespan thing.

      Ah, it's finally for sale somewhere! The $130 Newegg wants is a bit much, considering Gigabyte said it would retail for about fifty bucks though. So $430 for a solid-state 4GB drive, or $300 for a solid-state 30GB drive, but the performance and lifespan sucks. Unfortunately, neither is particularly useful. Seems to me that $5/GB x 30GB = $150, though.

      --
      How are sites slashdotted when nobody reads TFAs?
    24. Re:Gb or GB? by thisislee · · Score: 1

      hmmmm. I guess that could be true if astronomers are like every other profession and there are few brilliant ones and the rest pretty much really don't know the moon from the sun, beacause that's pretty much what I see here.

    25. Re:Gb or GB? by scbysnx · · Score: 2, Insightful

      whoever modded parent up hasn't got a clue. Not trying to insult parent but its wrong. I've seen plenty of people on /. use GB and Gb interchangebly.

    26. Re:Gb or GB? by Rickler · · Score: 1

      Is the article incorrect? If it's only 2gigs of storage that's been around for quiet some time.

      --

      The human race is artificial intelligence created using object orientated programming.
    27. Re:Gb or GB? by mrchaotica · · Score: 1
      new (much?) higher density
      These 2GB flash chips are only twice as big as the 1GB ones in the iPod Nano.
      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    28. Re:Gb or GB? by Anonymous Coward · · Score: 0

      From the article referenced in the story:

              Samsung will start producing 16 gigabit Nand Flash chips this year...

              Gartner estimates that 16GB Flash drives will cost from about $90...

       
      That looks about right. Each chip holds 2Gb of data, that puts the retail cost at about $10.00 a chip which is not unrealistic.
       
      Looking at it from your point that they confused Gb with GB you would have a single chip cost of $90 which is way too expensive.
       
      I don't think that the Slashdot crowd have any issues with knowing what 16Gb chips and 16GB drives are.

    29. Re:Gb or GB? by Propaganda13 · · Score: 3, Funny

      If it doesn't fit into a 4 step profit plan or Soviet Russia joke, I'm not really concerned with the technical aspects of anything on Slashdot.

      In that case, in Soviet Russia Gb is greater than GB.

    30. Re:Gb or GB? by Ethan+Allison · · Score: 1

      If it's $5 a GB, wouldn't that mean it'd be $150 for a 30GB? Correct me if I'm wrong. (No sarcasm intended, by the way)

    31. Re:Gb or GB? by Jozer99 · · Score: 1

      You misunderstand, I was WISHING for $5/GB flash. Currently, its about $70-$90/GB. I did a factor of two for the controller (which would have to be beefy for 30 flash chips, and manufaturing, and all the other various things that going into a product's sale price.

    32. Re:Gb or GB? by curtHendzell · · Score: 1

      Too true, too true.

      Good call.

      --
      -=Curtis=-
    33. Re:Gb or GB? by Shanep · · Score: 0

      The difference between gigabit and gigabyte needs to be explained on Slashdot about as much as the difference between the Moon and the Sun needs to be explained to astronomers.

      Now just you hang on there a dang second Mr. Byproduct. Slow down with all the scientamific techno mumbo jumbo. Moon? Sun? Give it to us slowly.

      --
      War crimes, torture, lies, illegal spying... Would someone give Bush a blowjob, already, so he can be impeached?
    34. Re:Gb or GB? by Anonymous Coward · · Score: 0

      I'ma firm believer in the philosophy of a ruling class, especially since I rule.

    35. Re:Gb or GB? by HolyCrapSCOsux · · Score: 0
      Indeed. The Mod Squad needs a few definitions themselves:
      Main Entry: insight
      Pronunciation: 'in-"sIt
      Function: noun
      1 : the power or act of seeing into a situation : PENETRATION
      2 : the act or result of apprehending the inner nature of things or of seeing intuitively
      Therefore the GP should be +5 Informative not +5 Insightful.
      --
      0xB315AA8D852DCD3F3DCA578FD2E0BF88
    36. Re:Gb or GB? by Belseth · · Score: 4, Informative

      There's other uses than computers for large flash drives. I'm getting ready to pick up a Panasonic HVX200. They use a P2 memory cards as their primary recording medium. For 1080 your only other option is external hard drives. It's about 1 gig a minute at 1080/60. That translates out at 8 minutes for the largest card availible the 8 gig P2 which uses 4 2gig cards. Right now the cards are running around $2,000 but they'll drop fast as capacity goes up. They really start getting interesting when you can get a 32 gig card for $500. Even in today's market it isn't a competitive price for a hard drive but for video use given the advantges it would be very attractive. Cameras will help get capacity up and prices down so may be one day they'll make sense for computer hard drives. Everytime some one says we don't need more memory another use is found and need goes up. Terrabytes will start maxing out need for most traditional uses though. The problem will start to be organizing files since in the terrabytes most people wouldn't need to delete files. Video and graphics people are the only ones that may never be happy. Storing a single full res feature would still take quite a few terrabytes to store so if you do it professionally or are simply a serious film fanatic there's no practical limit to the storage that could be used.

    37. Re:Gb or GB? by Shanep · · Score: 2, Informative

      Of course there is also that 'number of writes' issue.

      True. However they can be used practically as system disks. I've been using CF cards for diskless firewalls for more than a year now. With OpenBSD I use softdeps and noatime and I've had no problems. I know of others who have done the same for years.

      --
      War crimes, torture, lies, illegal spying... Would someone give Bush a blowjob, already, so he can be impeached?
    38. Re:Gb or GB? by Anonymous Coward · · Score: 0

      when talking about density, i'm pretty sure doubling in capacity is good enough to not be called "only".

    39. Re:Gb or GB? by Detritus · · Score: 1

      The design and organization of modern computer memory is strongly influenced by the base 2 numbering system. That isn't true for hard disks and many other devices. A hard disk doesn't care if the sector or word size is not a power of 2. At the hardware level, it just stores sequences of bits.

      --
      Mea navis aericumbens anguillis abundat
    40. Re:Gb or GB? by Anonymous Coward · · Score: 3, Interesting

      XP can be installed to use minimal writes to flash and spread writes out to avoid wear on specific sectors. This is standard for XP embedded development.

    41. Re:Gb or GB? by uberdave · · Score: 3, Informative

      The computer world use of K=1024, although perfectly logical from a binary addressing point of view, is a hack of the SI prefix notation. Back in the not so long ago, some marketing dweeb noticed that 64K was actually 65536 bytes, and started selling computers that had 65K instead of 64. "Lookie here, Brandine", the uninformed consumer would say, "This here machine's got isself a whole extrie K".

      The situation worsened when hard drives and RAM started into the megabyte realm. Is a megabye 1024*1024, 1024*1000, or 1000*1000? And if a gigabyte is a thousand megabytes does it mean 1024*1024*1024, 1024*1024*1000, or... well you get the picture.

      Now as computer memories grew, so did their communications speeds. The telecommunications industry has always measured information in bits, as opposed to bytes. Not constrained by having to address these bits with other bits (as RAM and ROM manufacturers are) they did not adopt the K=1024 "standard", and followed the usual K=1000 meaning. So for them, a 56KB/S channel, meant 56000 bits (not bytes) per second.

      So, with no standard for whether the b in Gb meant bits or bytes, or whether it meant 2^30 or 10^9, people started to get fed up. In the late 90's the IEC standards people got together and layed out a new standard (the "bible" one might say, if one were into puns). Lower case b is for bits. Upper case B is for bytes. Kilo (K)=1000, and Kibi (Ki)=1024.

      Of course, it will take a few years for the world to adopt these standards. Old warhorses like myself (who remembers when BASIC had line numbers) will still be calling things by the old names for rest of our lives. Those of you who have never seen a rotary phone, or 8-track, or have never known a time without blogs have it easy.

    42. Re:Gb or GB? by alexq · · Score: 1
      The difference between gigabit and gigabyte needs to be explained on Slashdot about as much as the difference between the Moon and the Sun needs to be explained to astronomers.

      one of them is closer - the brighter one.

    43. Re:Gb or GB? by Domini · · Score: 2, Informative

      Yup, saw that too...

      But then everyone (including you probably) also seems to confuse gibibyes and gigabytes anyway.

      Slightly off topic:
      It's similar to the markings on watches where the maker claims 100M water resistant, but this is a ploy, since the 100M does not mean 100m and the measurement only indcates 'safe to bath'. Most buyers don't know this and this confusion has also spread to other cheaper manufacturers...

      Grr. Know your SI units and you can't get fooled!

    44. Re:Gb or GB? by k8to · · Score: 2, Informative

      The redeeming fact is that the number of writes is now well over the number that people actually achieve in any normal application, including general purpose computing and even swap. Modern flash devices intended for general use are capable of distributing the phyiscal usage across the entire device, and the number of writes keeps slowly climbing orders of magnitude upwards.

      --
      -josh
    45. Re:Gb or GB? by nacturation · · Score: 4, Informative

      Samsung will start producing 16 gigabit Nand Flash chips this year...

      Gartner estimates that 16GB Flash drives will cost from about $90...


      Where's the conflict? Flash chip != flash drive. Flash drives can often comprise multiple chips. Let's say we stack 8 of those 16Gb chips into one drive. How big is the flash drive going to be?

      --
      Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.
    46. Re:Gb or GB? by nacturation · · Score: 2, Funny

      Also, I'd venture to say that most astronomers know the difference between Gb and GB and use them daily, but when was the last time most slashdotters were even exposed to the sun?

      --
      Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.
    47. Re:Gb or GB? by adrianmonk · · Score: 1
      in some applications this [cost] is actually not a factor.
      [ ... ]
      give me a 4gb flash drive and i'll be all over it!

      OK, if price is not actually a factor, and you really want a 4 GB flash drive, then by all means, have at it. They're only about $1200 to $1500 for 4GB flash drives (with an ATA interface).

    48. Re:Gb or GB? by iocat · · Score: 2, Funny
      I am out of the country and homesick. Slashdot is not being true to form, and thus clearly it falls to me, reluctantly, to post teh followng response:

      1) Confuse people about Gb vs GB
      2) Invent second step
      3) ???
      4) Profit.

      --

      Dude, I think I can see my house from here.

    49. Re:Gb or GB? by ton1c · · Score: 1

      if you had 1GB chips for $5 each, it would be $150 for a 30GB hdd not $300....

    50. Re:Gb or GB? by Obi-w00t · · Score: 1

      With hardware read-only switch. That would be a true safeguard against OS hacking - and yes I am aware of the difficulties of patching with new users who wouldn't know how to use the switch or would leave it off. On another note people have already made chips with Windows embedded.

    51. Re:Gb or GB? by famebait · · Score: 1

      How about a new one? In Taiwan, they breed fluorescent green gigabits. Hmmmmm.... nope, doesn't look like it's got much in the way of legs.

      --
      sudo ergo sum
    52. Re:Gb or GB? by somersault · · Score: 1

      you mean twice as dence, eg 100% more dence, which is quite a lot more if you ask me :p

      --
      which is totally what she said
    53. Re:Gb or GB? by Sparr0 · · Score: 0

      Looks like you already have a second step, 3) is the one in need of inventing.

    54. Re:Gb or GB? by somersault · · Score: 1

      when you say a 'single full res feature' you dont just mean a movie do you, because a full res movie can fit on a DVD. The more space people have, the less efficiently they use it, it's quite sickening! I think the same thing happens as processor speeds increase also, people write poorer quality algorithms..

      --
      which is totally what she said
    55. Re:Gb or GB? by Anonymous Coward · · Score: 0

      you could easily fit 60 or so GB into a 2.5" drive shaped device

      Wouldn't there be problems with heat?

    56. Re:Gb or GB? by Zakabog · · Score: 1

      And I installed XP Pro on a 4 gig partition, I have Photoshop, and Office both installed. You probably have a swap file on your drive taking up a gig or two, I moved mine over to another drive.

    57. Re:Gb or GB? by mythosaz · · Score: 2, Interesting

      Incorrect.

      XPe (embedded) [or XP a'le carte] specifically has support for non-writable and limited-writable OS partitions, and cam be engineered to fit nicely on modern DOCs.

      I've done several builds myself.

    58. Re:Gb or GB? by Lumpy · · Score: 1

      Or that slower than a hard drive issue.

      Laptops are slow enough in the hard drive already with 7200rpm and higher near impossible and the IDE bus being a very slow 66mhz single data rate.

      Why we dont have SATA in a laptop with SATA 7200 or 10,000 rpm drives for some real performance really boggles me.

      I would rather have a 20 gig high speed drive over a 100 gig slow as hell current drive any day.

      --
      Do not look at laser with remaining good eye.
    59. Re:Gb or GB? by Lumpy · · Score: 1

      Sorry but windows CAN. Go get yourself a copy of BartPE and behold the windows in a read only state loading into ram.

      Solid state system boot drive, crapload of ram to run OS and apps from. this would work very well and certianly would be darn fast.

      4Gb of ram would be plenty to house and run very fast the OS,Office and other apps in a ultra portable long life portable.

      --
      Do not look at laser with remaining good eye.
    60. Re:Gb or GB? by prionic6 · · Score: 1

      No, because the drive obviously needs other parts in addition to the memory chips, there is development / marketing cost and the company selling them wants to make a profit.

    61. Re:Gb or GB? by m50d · · Score: 1

      Yes, if you read his post again you'll see he's talking about recording in full HD at 60Hz. Personally I don't see the point, but for some people you can never get good enough.

      --
      I am trolling
    62. Re:Gb or GB? by Anonymous Coward · · Score: 0

      Interesting to note that in Russia the word "GB" (spelled "ge-be") is used to mean "gosbezopasnost'" (state security). Remember KGB? The K is Komitet (comittee), the GB... Shashdot should know better :) .

    63. Re:Gb or GB? by v1 · · Score: 1

      if your system can boot USB, you're already set. Get a SanDisk Cruzer Mini 4. I use mine daily, though not for booting from. They're still kinda pricey, but worth it I think. Got mine off eBay for $235.

      Gotta watch out on eBay though, I've talked with a couple people that have been or know people that have been scammed lately. They hack the format on 512mb flash drives to appear like they are 4gb and sell them. You don't know you've been duped until you try to reformat or run out of space earlier than expected. Don't buy a large flash drive from anyone without good feedback.

      --
      I work for the Department of Redundancy Department.
    64. Re:Gb or GB? by v1 · · Score: 1

      currently, the form factor for "usb thumb drives" allows for about four units of space, two on each side of the pcb inside the drive. flash drives usually have one, but sometimes two, controller chips on them, taking up two of those areas of space. That leaves two more spaces available. The largest thumb drives that I know of are 4GB, made with a pair of 2GB chips. So at least for thumb drives, the chip count is very important. (I suppose they could squeeze one more chip in if only one controller chip was needed, but digital storage tends to be done in powers of 2 and that doesn't include "3")

      --
      I work for the Department of Redundancy Department.
    65. Re:Gb or GB? by petermgreen · · Score: 1

      afaict pretty much all disk like flash does wear leveling in hardware below the filesystem anyway so doesn't require special filesystems.

      --
      note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
    66. Re:Gb or GB? by petermgreen · · Score: 1

      sectors have been 512 bytes (half a binary kilobyte) for a very very long time on nearly all disk like media (some types of cd and possiblly dvd being exceptions). and nearly all filesystems are built arround this assumption (generally using some sort of cluster which contains a power of two greater than 512 number of bytes and is fixed in size for any given partition)

      --
      note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
    67. Re:Gb or GB? by nmg196 · · Score: 1

      > IDE bus being a very slow 66mhz single data rate.

      This is still WAY more than any current laptop HDD requires.

      > Why we dont have SATA in a laptop with SATA 7200 or 10,000 rpm drives

      You don't need SATA in a laptop disk yet. They're far slower (due to the reduced size and thus slower platter edge speed) than a 3.5" drive so why bother upgrading the interfaces to SATA when normal ATA is easily fast enough? It wouldn't make any difference.

      > I would rather have a 20 gig high speed drive over a 100 gig slow as hell current drive any day.

      Um, a 100G drive will be FASTER than the 20gb equivalent in the same range due to higher data density. A bigger drive (bytes wise) is almost always faster than a smaller one, yet you're implying you expect them to be able to make the disk faster by giving you less space? How/why would that work exactly?

    68. Re:Gb or GB? by astralbat · · Score: 2, Informative
      You can get laptops with 7200 RPM disk drives - they are known as desktop replacements. AFAIK, SATA is being used in the latest models now, even for the 4200 RPM drives.

      Just don't expect their batteries to last long.

    69. Re:Gb or GB? by petermgreen · · Score: 1

      i'd imagine you could get a few more in if you used ics directly mounted on the board rather than large surface mount packages. you could also possiblly have two pcbs stacked especially if you considered a fairly fat stick acceptable.

      --
      note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
    70. Re:Gb or GB? by SCHecklerX · · Score: 1

      I was thinking 'throughput' at first, since throughput is typically shown in bps rather than Bps. For example, our typical FreeSWAN gateway has a throughput of around 52Mbps.

    71. Re:Gb or GB? by RicktheBrick · · Score: 1

      I do not understand why manufacturing of NAND memory is easier than manufacturing of ram memory. If they can do 16Gb of NAND than they should at least be able to manufacture 16Gb of ram memory as ram is still volatile and NAND is non-volatile. But I believe the biggest ram memory is still 2Gb and I do not know if that is mass produced.

    72. Re:Gb or GB? by shaka · · Score: 1

      give me a 4gb flash drive and i'll be all over it!
      Sure, here.

      They list it as $599, but it's on sale for about $265, at least where I live (Sweden).

      --
      :wq!
    73. Re:Gb or GB? by Anonymous Coward · · Score: 0

      I would mod you insightfull if I had any mod points

      (tong in cheek)

    74. Re:Gb or GB? by drasfr · · Score: 1

      There is never enough space... I guess I am one of those people. I only want top quality.

      Unfortunately for me, I am in videographer, as well as a photographer, and a video collector.

      When I work on photoshop on some photos. I always work from the original 16MP, 16bits photos from my camera, right there it is already 96MB. It isn't rare that when I save it in .psd that the photo is between 300 to 600MB when you add the layers and etc... Considering that when I do a photoshoot I might have 10 to 20 images I will rework on.... It takes up a lot of space.

      Then on the other end, I also do video editing. From mini-dv, it is about 12GB/hour. You can also fill up a harddrive very quickly with this. And it is not even the new hires cam you can get.

      I also used to like putting all my DVDs, uncompressed on my computer so I could have the best quality video/audio watching them on my projector.

      so see... i have 2TB of storage at home, and it is full... Backup is an headache. I can't wait to add another 3TB when I have the money to invest... (8x500GB drives in an external SATA array).

    75. Re:Gb or GB? by kabocox · · Score: 1

      and 100's of gigs into a device the size of a 3.5" drive. The problem is that these devices would cost astronomical ammounts.

      Some company needs to play with the idea and try to get it to work. At current costs it is unworkable, but if the manufacturer had to ramp up flash output by 100X to stick enough chips in a drive their cost might come down, and if could get 300 GB internal flashdrive for a laptop some people would buy even if it was expensive.

    76. Re:Gb or GB? by Anonymous Coward · · Score: 0

      Not true, a lot times the moon is seen in the day time

    77. Re:Gb or GB? by Anonymous Coward · · Score: 0

      5*30 == 300? Wow! Great math!

    78. Re:Gb or GB? by engagebot · · Score: 1

      What are you talking about? You've always got the overlords to fall back on...

      --
      Han shot first.
    79. Re:Gb or GB? by Novous · · Score: 1

      The parent makes a good point. Flash cards are very slow when compared to hard drives. An 80x flash card (the highest I've seen) is only 12 Mbps! To contrast that, Western Digital's Caviar drive lists 602 Mbits/s (Max).

      I'd only use one of those cards in special situations. Especially since flash also costs much more per megabyte.

    80. Re:Gb or GB? by dozer · · Score: 1

      [watches] where the maker claims 100M water resistant, but this is a ploy, since the 100M does not mean 100m...

      Since "M" is not an SI unit, I don't see how "knowing your SI units" will help in any way. And, after doing some googling, I think you're wrong. In every watch ad I can find "100M" means "100 meters" even though "M" is capitalized. Care to provide a link supporting your statement?

    81. Re:Gb or GB? by Detritus · · Score: 1

      There are exceptions, like IBM mainframe DASD. If you are a large enough customer, you can order drives with non-standard parameters.

      --
      Mea navis aericumbens anguillis abundat
    82. Re:Gb or GB? by jqpublic13 · · Score: 1

      What am I missing...? $5/GB = $150/30GB, not $300... and if current market conditions hold true, the larger the drive, the *cheaper* the cost/size, not the other way around...

      --
      Non calor sed umor est qui nobis incommodat.
    83. Re:Gb or GB? by The+Ultimate+Fartkno · · Score: 1

      A fluorescent green pig with no legs?

      I shall pet him and love him and name him "Porky Pickle."

    84. Re:Gb or GB? by poot_rootbeer · · Score: 1

      one 8Gb chip is the equivalent of 1GB.

      But when they say 8 gigabits, do they mean 2^33 bits (8,589,934,592), or 8x10^9 bits (8,000,000,000)? At that magnitude, the difference between base 2 and base 10 is not trivial.

      The first defintion SHOULD actually be referred to as gibibits, but that nomenclature is not yet universal. OS developers, I'm looking in your general direction here.

    85. Re:Gb or GB? by poot_rootbeer · · Score: 1

      Gartner estimates that 16GB Flash drives will cost from about $90...

      Where's the conflict? Flash chip != flash drive. Flash drives can often comprise multiple chips.


      It's a typo. Unless Gartner actually expects that 128-gigabit drives are going to sell at the same price 16-gigabit Flash drives currently do.

      Maybe by the year 2010 they will.

    86. Re:Gb or GB? by undercanopy · · Score: 2, Informative

      uhh... i think you mean 12MBytes/s, but they're getting even faster than that.

      I don't know what kind of flash drive you have, but mine runs a hell of a lot faster than usb1.1 speeds. I haven't tun hdtach on it but it but i just copied 87 megs of mp3s onto it in ~6 seconds, that's 14.5 MBytes/sec, or 116Mbits/s

      looking back up i see you said flash card... dunno fi you think it matters, but my 512MB SD card is narly as fast.

      --
      -- D-23994, Muff#2613
    87. Re:Gb or GB? by Surt · · Score: 1

      On the other hand, a watch manufacturer is stupid to be unclear in their feature advertising: if the product fails under conditions the user expects it not to, you've surely lost a customer forever, and the replacement watch is not going to be built by you.

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    88. Re:Gb or GB? by generationxyu · · Score: 1

      Stop posting.

      --
      I mod down pyramid schemes in sigs.
    89. Re:Gb or GB? by tetabiate · · Score: 1

      A friend who is a mathematician told me jokingly sometime ago that although from the topological point ov view mugs cannot be distinguished from donuts, only in extremely rare situations you see mathematicians serving coffee in donuts and eating mugs.

        - Ah! Friday afternoon. Thinking about going out and have beer, pizza and... you know,

    90. Re:Gb or GB? by rwise2112 · · Score: 1

      So, size does matter! Oh crap!

      --

      "For every expert, there is an equal and opposite expert"
    91. Re:Gb or GB? by GmAz · · Score: 1

      I totally agree with you. If I had a four or five gigabyte flash drive, I would put my OS on it and nothing else. When I see a cheap USB2 flash drive with that capacity, I might try it since you can boot to USB2 drives if your motherboard supports it (which mine does).

      --
      Click Click Bloody Click PANCAKES!
    92. Re:Gb or GB? by mrchaotica · · Score: 1

      Ever heard of Moore's Law? Microchips double in density all the time. In fact, I would have been more surprised if they hadn't come out with these sometime soon...

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    93. Re:Gb or GB? by somersault · · Score: 1

      that's not how Moore's law is defined from wikipedia: "empirical observation that at our rate of technological development, the complexity of an integrated circuit, with respect to minimum component cost, will double in about 18 months." ... "Under the assumption that chip "complexity" is proportional to the number of transistors, regardless of what they do, the law has largely held the test of time to date." Nothing to do with chip density. Making chips twice as dence seems quite impressive to me.

      --
      which is totally what she said
    94. Re:Gb or GB? by aonaran · · Score: 1

      Why not? it was only a year ago that $90 was the going rate for a 256Mbyte usb key. ... we've quadrupled the storage for that price in a year, why would we not be able to do the same this year??

    95. Re:Gb or GB? by Anonymous Coward · · Score: 0

      I wonder if we'll ever have a microchip event horizon.

    96. Re:Gb or GB? by moonbender · · Score: 1

      You're right, they're faster than the grandparent makes them out to be. But 16 MB/s is still only a third to a fifth of a fairly ordinary desktop HD.

      --
      Switch back to Slashdot's D1 system.
    97. Re:Gb or GB? by mrchaotica · · Score: 1

      No, but we will have a point at which it's impossible to make transistors any smaller. In fact, we're almost at that point now with CPUs -- that's why Intel, AMD, etc. are having so much trouble making chips faster and instead are turning to parallelism.

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    98. Re:Gb or GB? by default+luser · · Score: 1

      I do not understand why manufacturing of NAND memory is easier than manufacturing of ram memory. If they can do 16Gb of NAND than they should at least be able to manufacture 16Gb of ram memory as ram is still volatile and NAND is non-volatile. But I believe the biggest ram memory is still 2Gb and I do not know if that is mass produced.

      Good question. Here is my take:

      Memory makers typically push their latest process technology using NAND Flash because NAND Flash is a heavy growth market, and needs to be as small and low-power as possible. This means that they don't yet encounter unpredictable downturns. For example, Samsung was already using a 73 nm process on their NAND Flash as they recently moved their DRAM to 90 nm. Basically, when you combine the need for very small size and low-power, plus the huge growth market, the higher the density, the better.

      NAND is also not hard to make in high densities because most manufacturers have developed ways to make one cell (two transistors) hold two bits of data. This is as-opposed to DRAM, which stores each bit using a transistor and a capacitor. Depending on the size of the capacitor, you should be able to get better density from a NAND Flash chip than DRAM.

      DRAM cannot hope to reflect Flash's growth, and DRAM is also more volatile because its growth is directly attached to another volatile market: computers. To make this clear: people may buy multiple flash devices for each computer they own (including new AND existing computers), but they will typically only buy one or two sticks of DRAM, and typically only on new computers. In addition, the capacity of memory makers can sell is limited by market demand, so higher-density memory becomes wasteful.

      While the potential growth for NAND is still "limitless", the mainstream DRAM market is very limited, and is dangerously tied to new PC sales. Thus, makers use tried-and-true processes with very low defect rates, and carefully select density based on a a balance of cost and market demand for larger-capacity DIMMs.

      You might point to cutting-edge, high-end parts like GDDR3 on graphics cards as a high-growth market, but compared to sales of DIMMs, sales of high-end video card memory are small...they are also very low density parts, mostly to make the die size smaller and thus reduce the defect rate.

      --

      Man is the animal that laughs.
      And occasionally whores for Karma.

    99. Re:Gb or GB? by undercanopy · · Score: 1

      true, though that was write speed, reads are even faster... still a fraction, but a more respectable one.

      One thing that should be taken into account, though, is that this is all well and good for contiguous files. Throw in some fragmentation or just multiple sparsely located files and the aggregate seek times will very quickly close that gap.

      This goes back to the rambus/dram argument of throughput vs latency... and in this case we're talking about latency differences measured in several orders of magnitude, that means everything for random data access.

      --
      -- D-23994, Muff#2613
    100. Re:Gb or GB? by TClevenger · · Score: 1

      Don't forget that hard drive manufacturers were the ones to change things on us and start this mess. Back in the day, a 20MB ST-225 hard drive had well over 20,000,000 bytes (usually over 21,000,000 bytes, depending on bad sector mapping at the factory.) It wasn't until multiple-gigabyte drives that manufacturers started putting in the "1 GB=1,000,000,000 bytes" nonsense to inflate their drive sizes.

    101. Re:Gb or GB? by rhandir · · Score: 1

      Try the Windows Powertoys TweakUI.

      Makes it easy to move not just /My_Documents/ to another drive/folder, but also /Program_Files/.

      You want to move /Program_Files/ to your D: drive (big IDE drive.) Photoshop will thank you btw, it creates its own magic swap file, and prefers it to not be on the same disk as window's swap.*

      If you want to trim up windows, investigate nLite (http://www.nliteos.com/nlite.html). It is a menu-driven program for making windows install cd's, which you can use to tweak almost every parameter, down to which services are disabled/manual when it is installed. It will slipstream in hotfixes, and you can in fact slipstream in applications, should you wish to do so. (The last requires hacking the installers for those apps with a hex editor. Not for the fainthearted.) nLite does NOT remove the requirement for authentication or for product keys, so no free lunch to you warez folk.

      -r.



      *allegedly making window's swap a constant size also improves performance/prevents fragmentation. To fix it on an existing install, create a new swapfile on D:, remove swap from C:, reboot, defrag C:, create a new swapfile on C:, get rid of swap on D: Given a constant size, it shouldn't spill all over you harddisk. See WindowsXP Annoyances, 3rd ed.

      If you are fanatical about performance, move the applications to another HD on a different ide channel (not slave) or onto a scsi or SATA channel that is separate from the OS. If you are really fanatical, put the os and swapfile on a SATA raid 0 configuration, and separte data and programs as suggested. None of these last solutions have been tested by me, your milage may vary.

    102. Re:Gb or GB? by default+luser · · Score: 1

      Exactly, all Creative Labs Muvo flash players (and likely the N200 series) use this concept. They have two PCBs bridged by a modular connector. The main PCB has the logic, USB connector and buttons. The other PCB has the memory and LCD. So, they can take advantage of this and produce multiple memory PCBs, and use the same main PCB for all device variations.

      Have a look here for deconstruction of the Muvo 512MB

      --

      Man is the animal that laughs.
      And occasionally whores for Karma.

    103. Re:Gb or GB? by TheRaven64 · · Score: 1

      Flash completely wipes the floor with hard drives when it comes to seek times, which are a huge performance killer in most disk-intensive operations. When it comes to sustained transfer rate, this is quite easy to fix. Disk reads are done in blocks of 4KB. Flash reads are done serially. You can put up to 4096 flash devices in a single module, reading singly byte from each for every disk block. You can put 32768 in if you read one bit from each. At this speed, they would wipe the floor with hard drives in terms of transfer rate too. The real problem is one of cost and space. Rather than adding chips, manufacturers are more likely to simply increase the address width of each chip (which already contains a number of individual banks).

      --
      I am TheRaven on Soylent News
    104. Re:Gb or GB? by buck_wild · · Score: 1

      I think that you're correct in your reference to /. readers, but when the article itself refers to both Gb and Gb for the same product, people want to know which it is before they get excited about it.

      --
      If all you have is a hammer, everything looks like a nail.
    105. Re:Gb or GB? by Anonymous Coward · · Score: 0

      in Soviet Russia Gb is greater than GB.

      In Soviet Russia it's KGB

      Which brings an interesting question, is 1KGB equal to 1TB?

    106. Re:Gb or GB? by petermgreen · · Score: 1

      one thing i noticed in those pics was a BGA chip. BGA chips have a HUGE connection density BUT afaict they have to be mounted on top of a grid of vias preventing anything being placed on the other side of the board opposite them.

      i wonder if this was why they wen't for the two board structure

      --
      note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
    107. Re:Gb or GB? by Audacious · · Score: 1

      This has been my question. Why not do these things like the old memory cards for the Apple //e et al systems?

      You have a controller that decides which bank of memory to store the information. There are several banks of memory, the controller makes it all look like one continuous amount of memory. I just looked at PriceWatch for USB flash memory units and they show a 1GB at $55.00 and a 512MB at $30.00.

      So ten 512MB = 5.12GB at $300.00, but if these were manufactured like this I would think the cost would drop to around $150.00 or maybe $200.00 for the unit.

      Or ten 1GB = 10GB at $550.00 (or maybe $300.00 - $400.00).

      I would think that whatever type of controller they are already using should be expandable to handle the larger sizes.

      I'm also wondering about the heat problem. Does anyone know if the larger sized drives have a heat problem? Are any of them using micro fans or something like that? Can you post if you know of anything along these lines? TIA!

      --
      Someone put a black hole in my pocket and now I'm broke. :-)
    108. Re:Gb or GB? by default+luser · · Score: 1

      No, they don't.

      Take a closer look at those pictures, starting with the one that is a closeup of the Sigmatel chipset. Below that picture is another picture with two PCBs. The PCB on top in the picture is the one with the Sigmatel chipset (that chipset is on the back side, not shown in that picture).

      Do you see any large collection of vias on the backside? I certainly don't. In fact, that's the most component-packed side of either of the two PCBs...it's really crowded.

      BGA chips aren't much different from SIPP-style packages...they still connect to solder points on the surface of the board. You just can't see them.

      --

      Man is the animal that laughs.
      And occasionally whores for Karma.

    109. Re:Gb or GB? by SimonInOz · · Score: 1

      Surely astronomers would be exposed to the sun even less than Slashdotters?

      Nah, not possible.

      --
      "Cats like plain crisps"
    110. Re:Gb or GB? by devilspgd · · Score: 1

      > I would rather have a 20 gig high speed drive over a 100 gig slow as hell current drive any day.

      Um, a 100G drive will be FASTER than the 20gb equivalent in the same range due to higher data density. A bigger drive (bytes wise) is almost always faster than a smaller one, yet you're implying you expect them to be able to make the disk faster by giving you less space? How/why would that work exactly?

      You misunderstand -- He's not suggesting the 20gig drive would be faster, just that given the options, he would take a faster 20gig drive over a slower 100gig drive.

      --
      Give a man a fish, he'll eat for a day, but teach a man to phish...
    111. Re:Gb or GB? by Domini · · Score: 1

      It has become to represent meters due to mass misunderstanding since it used to be meters when it was still called 'waterproof'. A lot of wrongs still don't make a right. The term 'rater resistant' does not *literally* represent the depth of general use (unlike 'waterproof'), but it is more like a placeholder for another different but similar measurement. Manufacturers prefer this symbolic substitution since it serves their purpose. It's like saying that my car can use up 1/3 of it's tank in 250km and it reaches optimal temperature in 1 hour, and thus my engine has a rating of 250KM/H. This is now my own custom rating standard. KM still sorta stands for km and is basically the same, but still misleading. (Perhaps this is a poor example, but you get my drift, right?)

      100M (or 100m or 100 meters) goes a long way to mislead about the depth of use to the uninformed as this really means: A watch marked with 100 Meters should be able to survive surface swimming.

      Knowing your SI units would lead you to question: why the discrepancy? Even on diver's forums I find people who discuss this rating, but all of them are under a false impression.

    112. Re:Gb or GB? by davidrb84 · · Score: 1

      OH dear, i'm 22 and we had a rotary phone in our spare room when i was little, and used to code in BASIC writing thing like 10 print "teacher smells" 20 for i = 1 to 1000 () 30 goto 10 forgive that code, it's been a long while since i've done any BASIC ..... bring back line numbers! (and yeah we were using pretty poor acorns at school where, for i = 1 to 1000 would actually give a pause in automation, no idea why it didnt parse that out when it compiled as it usually contained no code! aaawwww the old days of 1990s

    113. Re:Gb or GB? by jsiren · · Score: 1
      Hrm. There may indeed be some confusion. One chip has a capacity of sixteen gigabits (16 Gb). Assuming a byte size of eight bits, a sixteen gigabyte (16 GB) drive contains eight 16 Gb chips, as implied by the document you quoted.

      A complete drive also contains an interface or bus controller chip if mandated by the spec, additional electronics as required to create a functional device, a circuit board, a connector to interface with the rest of the world, and a protective case to avoid interfacing with the unfriendly parts of the world.

      --js--

      --
      Usage: km/h for speed (kilometers per hour); kph for very slow impulses (kilopond hours).
  2. One Thought... by Architect_sasyr · · Score: 3, Interesting

    Burnout.

    What is the burnout like???

    --
    Me failed English...
    FreeBSD over Linux. If my comments seem odd, this may explain...
    1. Re:One Thought... by yobjob · · Score: 0

      Why worry about burnout? You only have to install Windows on the drive once, right? Oh wait..

    2. Re:One Thought... by BadassJesus · · Score: 5, Informative

      Flashdisks are much more reliable then any conventional harddrive. They claim >5,000,000 write/erase cycles and unlimited reads. Unites States defence department is using them for reliability issue alone.

      M-Systems (top flash disk producer) states this:
      (copied from the website)

      Top Reliability & Endurance
      ** 99.999% reliability
      ** >1,400,000 hours of actual (in the field) MTBF
      ** Embedded EDC/ECC, based on BCH Algorithm
      ** Data integrity under power-cycling
      ** TrueFFS® technology: bad blocks mapping-out and dynamic wear-leveling algorithms
      ** >5,000,000 Write/Erase cycles; Read unlimited
      5-year warranty


      Source link:
      http://www.m-systems.com/site/en-US/Products/IDESC SIFFD/IDESCSIFFD

    3. Re:One Thought... by AuMatar · · Score: 1, Interesting

      Having worked on flash memory in embedded projects, I know of no product with that long a life cycle. Additionally, as size increases burnin problems tendto increase as well.

      And that MTBF you site would be 160 years. So I see no way in hell that can be in the field, flash didn't exist 160 years ago. That throws all the other numbers into the trash.

      --
      I still have more fans than freaks. WTF is wrong with you people?
    4. Re:One Thought... by nexthec · · Score: 1

      You need to do more research on how MTBF is usually found

    5. Re:One Thought... by cdrudge · · Score: 5, Informative

      MTBF isn't absolute. It's a statistical estimate. A hard drive may have a 500,000 hour MTBF. That particular model of drive wasn't tested for 57 years to see if it failed.

      Any type of failure rate is also representive of the collection of all products being tested, not a single one.

      Read the Failure Rate Wiki entry for more information.

    6. Re:One Thought... by Brad1138 · · Score: 2, Funny

      "Unites States defence department is using them for reliability issue alone."

      Boy I'm sold!

      --
      If you could reason with religious people, there would be no religious people
    7. Re:One Thought... by grcumb · · Score: 3, Funny

      "** >1,400,000 hours of actual (in the field) MTBF "

      Hmmm. 1.4 million divide by 24... that's, uh, carry the one... about 58,333 days. Which would be, uh... ah, ignore the leap years... Almost 160 years. That means they've been testing this hardware since before the Civil War!

      Wow, now is that dedication or what? Where do I buy me one of these babies?

      --
      Crumb's Corollary: Never bring a knife to a bun fight.
    8. Re:One Thought... by Anonymous Coward · · Score: 0

      ** >5,000,000 Write/Erase cycles; Read unlimited

      Bullshit! I work on flash memory designs in embedded control. Current flash tech is just appraoching 1E6 write/erase cycles. Something stinks here.

    9. Re:One Thought... by MacroRex · · Score: 1

      You don't want their chips, you want the time machine they've obviously been using to test the chips.

    10. Re:One Thought... by jolyonr · · Score: 1

      That's assuming the company making your flash drive doesn't try to cheat you by using unsuitable chips - as happened to me:

      Why Goodmans USB keys should not be trusted

      And they're still refusing to answer any of my emails on this subject!

      Jolyon

      --


      Please read my Canon EOS tech blog at http://www.everyothershot.com
    11. Re:One Thought... by gnasher719 · · Score: 4, Informative

      '' MTBF isn't absolute. It's a statistical estimate. A hard drive may have a 500,000 hour MTBF. That particular model of drive wasn't tested for 57 years to see if it failed. ''

      Also important: Products like harddisk have a limited life. That harddisk with 500,000 hour MTBF will wear out after five years or 50,000 hours; no way will it last 500,000 hours. The MTBF only means: If you buy 500 harddisks and run them for 1000 hours, you can expect one to fail.

    12. Re:One Thought... by _mythdraug_ · · Score: 1

      MTBF != Lifespan of a single unit.

      If I have 1000 units all started at the same time, one failure in 58.3333 days from those 1000 units will result in the same MTBF as quoted in the article. If you expand the base to 10,000 units; a single failure every 5.8333 days will again get you the same MTBF.

    13. Re:One Thought... by Anonymous Coward · · Score: 0

      I wonder how a Raid type system wold work for these drives. Chip 12 has just burned out, please replace at next conveneance.

    14. Re:One Thought... by Surt · · Score: 1

      Much like the 3,000,000 MTBF figures offered on some hard drives, these numbers are figured out by running a large number of test devices in parallel for one or more years, finding out how many fail, and using statistics to figure out what the average failure time is therefore going to be. And before anyone asks, yes, the statistics also account for accelerating wearout over time.

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    15. Re:One Thought... by Godeke · · Score: 1

      *BZZZT* If the *MEAN* time between failures is 500,000 hours it indicates that given a sampling of devices you will find the arithmetic mean lifespan to be 500,000 hours. If you expect 1 drive in 500 to fail in 1000 hours, then you must have the expectation that at the other end you have a near immortal drive. Considering that the failure curve for most electronics follows the "burning, lifespan, fatigue" curve (lots of failures up front which can be tested to remove prior to market, a long period with low failure rates and finally a spike when the materials give up the ghost at "end of life") if they are doing adequate burn in you can shift the MTBF experienced by customers to the far end quite well.

      While I would agree that 57 years seems a bit optimistic, we *are* talking about a device with no moving parts, known rewrite schedule and error correcting algorithms. Also, realize that the rewrite schedule is doing wear balancing, reliability testing and retirement of potentially unreliable blocks. In this situation, an *expected* MTBF of 57 years may be possible, but you may exhaust your block rewrites before then... I suspect that isn't considered a "failure" since the device operates to spec and the data is still readable.

      --
      Sig under construction since 1998.
    16. Re:One Thought... by sankyuu · · Score: 1

      Interesting insight from godeke. Here is my commentary:

      *BZZZT* If the *MEAN* time between failures is 500,000 hours it indicates that given a sampling of devices you will find the arithmetic mean lifespan to be 500,000 hours.

      As is obvious, the GP is likely using the formula:

      drives * failure rate * time = number of failures
      an average drive will fail in 500,000 hours:
      1 drive * (1 failure / 5e5 hours) * 5e5 hours = 1 failure
      if 500 drives run for 1000 hours, how many failures might there be?
      500 drives * (1 failure / 5e5 hours) * 1e3 hours = 1 failure

      However, this assumes that the likelihood of failure of each drive is evenly distributed within those 500,000 hours (curve is flat) -- even if drives were bought at the same time, something like 1 drive dies every 1000 hours.

      If you expect 1 drive in 500 to fail in 1000 hours, then you must have the expectation that at the other end you have a near immortal drive.

      What the GP pointed out is that actual drives tend to fail towards the end of their lifetimes, so that the death curve starts low and then shoots up later. They'll start healthy, then die en masse when they get old.

      The trouble with averages is that they make the curves look flat. Interesting to keep this in mind when shopping for equipment...

    17. Re:One Thought... by Godeke · · Score: 1

      I wish the expected failure curves were documented because of that exact problem: averages can mask a large standard of deviation or a small one. The post I buzzed *did* assume a massive one (a "flat" curve) but I don't think I have ever seen a flat failure curve. As described, you have manufacturing defects (the spike of failures in the short run) the useful life (a flat period where failures are rare) and finally the fatigue point where the curve spikes again.

      With a documented failure curve, a manufacturer could assure me that they have burned a product in sufficiently that I can expect to be using the device during the useful life, and thus allow hot spare planning to be much easier. However, the reality is that I have received "bad batches" of devices which all fail in the short run because of a defect, leaving my hot spares insufficient to handle the failure rate. Nothing reassures people *more* than having to scramble for more units because the devices are still in the burn in period when we get them.

      --
      Sig under construction since 1998.
  3. gigawhat? by Daniel+Dvorkin · · Score: 4, Informative

    Seems like they're playing fast and loose with capacity: "will start producing 16 gigabit Nand Flash chips this year" vs. " currently in products such as USB drives and digital cameras in capacities of up to 8GB." Correct me if I'm wrong, but doesn't 16 gigabits = 2GB?

    --
    The correlation between ignorance of statistics and using "correlation is not causation" as an argument is close to 1.
    1. Re:gigawhat? by kf6auf · · Score: 2, Informative

      It is my understanding that individual chips currently max out at 8 Gb; so t have an 8GB capacity right now you need 8 chips in your USB drive/camera/whatever. "Tomorrow" you will only need 4, meaning it should cost about half as much once the fixed costs are paid for.

    2. Re:gigawhat? by Amouth · · Score: 1

      the trick is the foot print of one chip..

      the 2/4/8 GB things you see now stack multi chips.. It is the same as HDD and areal density on the platters the larger the drive the denser it gets..

      --
      '...if only "Jumping to a Conclusion" was an event in the Olympics.'
    3. Re:gigawhat? by Mantus · · Score: 3, Informative

      Your problem is the assumption that a device may only use one chip. There are 12GB (~$7300 USD) CF cards available and they use chips with less capacity than 16Gb. This will allow higher capacity/cheaper devices.

    4. Re:gigawhat? by Skal+Tura · · Score: 1

      Let's not forget that 2Gb is twice as much as before on a chip like that, and they are really small, consume very little electricity, and you can slam like 100 of these on the same space as 1 regular hdd, altho the costs won't be anywhere near the same pricing.

      And look at the price, cheaper than what i've seen 1Gb Flash chips selling!

      2Gb(Gigabytes) doesn't sound much, but having 2Gb in my digital camera... Huh, i can go REALLY wild with it, and it'll take ages to fill ;)
      "Now, where are you GIrls! ... " >;D

    5. Re:gigawhat? by Godman · · Score: 1

      You just use 4 of them....

      --
      I have this really funny quote that I like to put here. Unfortunately, there's this really annoying thing called a char
  4. A Correction by iluvcapra · · Score: 1

    I think the post needs a correction. It should probably read:

    maybe even edging out hard drives in some PAINFULLY SLOW products in the next few years
    --
    Don't blame me, I voted for Baltar.
    1. Re:A Correction by Fishead · · Score: 3, Informative

      Ummm.... prolly not.

      I don't know any specifics (but neither did your post, so we are even) but I build computer controlled devices that need to work in a fairly high vibration environment. Our current product runs off Win2K, and boots relatively quickly off a 2GB solid state Laptop size HDD.

      We are building a new product that will be running Linux off a 256MB CF card. We are not quite done development, but it seems to run OK. It isn't working that hard though, just polling a USB control panel and outputting control commands based on what the user wants to do with a small amount of very simple graphics.

      (disclaimer: I know very little about OS's and software. I am mostly solder jockey, circuit design, system installer, and a little bit of sales. I have been struggling to get my MythTV box working for over a year now :-)

    2. Re:A Correction by DAldredge · · Score: 1

      My Lexar Lighting 512MB thumb drive is rather fast (22MB - 25MB per second). While not quite up to moving platter speed it is still a marked improvement of the thumb drives of the past 2 - 3 years.

  5. Re:I'm not surprised by Anonymous Coward · · Score: 0

    I am not interested in a notebook with a 2GB storage capacity. *Maybe* a handheld device, but not a bulky notebook. 1996 was ten years ago, let's get real.

  6. Vista Won't Fit? by students · · Score: 2, Interesting

    This article indicates that Windows Vista won't fit on a 16 Gigabit drive? And I thought MS was disk space hungry today. I used to use a gigabyte partition for Mandrake Linux - including applications and configuration - but not user data. Windows XP needs a gigabyte without applications. MS is crazy.

    1. Re:Vista Won't Fit? by msully4321 · · Score: 1

      16 gigabit = 2 gigabytes.
      That's not unreasonable at all (these days).

      --
      Slashdot: You will never find a more wretched hive of spam and zealotry. We must be cautious.
    2. Re:Vista Won't Fit? by students · · Score: 0, Troll

      Yes it is. The size of software, particularly software that has no real additional features, should not grow exponentially.

    3. Re:Vista Won't Fit? by SavvyPlayer · · Score: 1

      But Vista is basically Doom3-meets-OS-meets-corporate-machine. The Doom3 part of Vista craves many many megabytes of disk space.

    4. Re:Vista Won't Fit? by Skal+Tura · · Score: 1

      16 Gigabits is 2 Gigabytes, 2 Gigabytes as in a bit under 3 CDs capacity (700mb ones).
      So Vista won't fit into one of those, but the price is good, and they are small.

    5. Re:Vista Won't Fit? by izomiac · · Score: 1

      If you move the "Documents and Settings" and "Program Files" folders and the page file to other partitions then it might be possible to trim an XP install to under 2 GB. I do that on my XP installation (moving those folders is a pain, but possible), and it usually stays at about 3 GB used space. If you disable hibernation support and made sure no Windows Update files were there then you might be able to fit an XP install on a 16 Gb drive. If you started moving .dll files then you most certainly could. It'd be much easier to just use a 4 GB card though.

    6. Re:Vista Won't Fit? by Plaid+Phantom · · Score: 1

      Yes, and the corporate machine part craves your soul.

      --
      All comments are properties and trademarks of the voices in my head. Not like I'm gonna claim them.
    7. Re:Vista Won't Fit? by Xtravar · · Score: 1

      I used to go to all that hassle of moving those folders to different drives. I got my Windows partition down to less than 1GB after lots of tweaking. In fact, I think I got it down to around 700MB after a lot of compressing and deleting useless files and such.

      But recovery applications then couldn't find those folders.

      And every couple years, you need to redo the whole thing anyway. That's just life.

      And stupid applications still install into C:\WINDOWS and C:\WINDOWS\SYSTEM32

      So, I decided that it was easier to simply install linux, since it isn't a pain in the ass to partition the system the way I want.

      --
      Buckle your ROFL belt, we're in for some LOLs.
    8. Re:Vista Won't Fit? by Anonymous Coward · · Score: 0

      In a completely unrelated note, a friend tried to install Ubuntu onto an old computer last week that has a 2 GB disk, and it ran out of space during the install. Sure, you can play with it to install without everything you don't need, which is what he did, but it was a bit of a shock.

    9. Re:Vista Won't Fit? by somersault · · Score: 1

      seems pretty unreasonable to me if you actually read the parent post. WTF do they need 2GB for?!? Just because we have larger drives doesnt mean that MS should be allowed to bloat their OS even more.. also dont forget that the Windows folder tends to increase in size with crap over time as well *goes to check* at work here my machine somehow has 3.84 GB in the Windows (XP) folder - and to think that user profiles used to be kept in the Windows folder too.. there is another 350mb in my 'documents and settings' folder (no I dont use 'my documents', nor do I hold anything but shortcuts on my desktop). If linux and a whole office suite and other apps fit on a single CD, or even I think on a 128MB, or at least 256MB memory key, then there is no excuse for using 2GB for the OS alone... how innefficient do you think the code must be? It's not like Windows Vista is a game that has FMV sequences or hundreds of MB of textures and sound effects (though if the 2GB does include lots of desktop themes then that helps to explain thigs a bit)

      --
      which is totally what she said
    10. Re:Vista Won't Fit? by dhasenan · · Score: 1

      Let's see. Most desktop Linux distributions take up between 1 and 3GB, I believe; Ubuntu takes about 1.8GB but requires a root partition of 4GB. RHEL takes 1.8GB as well. (Neither of these feature development tools, though--even getting a sane build environment with Ubuntu takes half an hour or so.) WinXP is slightly heavier and gives you less; it's perhaps 2.5GB, but doesn't have such arcana as a Lisp interpreter built into Notepad. Instead, it has a number of poorly understood services which, disabled, may crash your computer or make it run slightly faster.

      However, I've seen custom Windows installations that take up less than 500MB. It's a matter of ease--with Linux, someone else has done this and released their distribution, even has a website for it; with Windows, someone else has done this and released the results on BitTorrent without even a forum for support.

    11. Re:Vista Won't Fit? by somersault · · Score: 1

      Yes, but those 'windows installations' dont have Office etc, though they do have wordpad which I actually like (compared to notepad at least).

      With the windows thing, making a custom installation would not be as simple as for linux, and almost certainly it's illegal anyway to release that custom version so meh.

      Windows also doesnt feature developement tools as part of the base installation, and in fact the base install of Windows is pretty useless unless you want to play solitaire, while you can get base installs of Linux running on small flashdrives with OpenOffice etc. I just dont like the mindset that people have of accepting that Windows really has to take up so much space and be so bloated - if it were better designed and coded then it would be much more efficient and reliable.

      I've been using Ubuntu at home for a couple of months now, and have on a machine at work (though I mostly leave the work machine in windows to access things more simply over the domain). Maybe the full install takes over a Gb, but you can run it from a live CD with OpenOffice, GIMP, etc, I'd like to see someone try to squeeze Vista, Office Vista and Photoshop onto a live CD and run as reliably.. Anyway

      --
      which is totally what she said
  7. Re:I'm not surprised by Jotaigna · · Score: 3, Informative

    i dont know about long lifed, since a flash memory card has a limited number of writes anyway. This is because each one of the memory holding transistors has an extra insulating layer between the gate and the emitter, so electrons are "forced" throug, trapping the bit and therefore the data. There is only so many times (in the 10s of thousands) you can do this, and then is toast. If you use it as a hard drive with crappy memory paging, it will die soon.

    --
    "The quality of life is inversely proportional to the number of keys on your keyring."
  8. Flash is ready even now by Anonymous Coward · · Score: 5, Insightful

    There are linux distros that happily run on flash. Damn Small Linux comes first to mind. It's possible, in fact many people have done it, to build a computer with no hard drive; just flash.

    The current problem is that you get only a limited number of writes to flash. TFA doesn't mention that. It is a problem but not an insurmountable one.

    1. Re:Flash is ready even now by Anonymous Coward · · Score: 1, Interesting

      Hold the OS/apps in the NAND device, but *run* it from one of those PCI card persistant RAM (from a battery) stack sets from AMD. Zillion writes from the RAM stack, problem solved. And even if the battery goes kerflooey, you still have the original in the NAND drive. It only needs to read once on boot then, and you can write to capacity on the RAM stack, then transfer it to an optical disk for longer term storage. Perfectly good enough for a diskless laptop or PDA thingee as long as you don't want to hold a hundred thousand videos, etc, but that is what your cheap throw away vid3oggpod is for..

    2. Re:Flash is ready even now by earnest+murderer · · Score: 1

      It depends on who made it and how much you want to pay. Twenty years ago NAND was good for 100,000 write cycles. Now most manufacturers claim about 1 million for comodity chips and bandwidth around 100/50 MB read/write per second.

      With some memory management changes (swap is a crutch anyway ;) it shouldn't be an issue for the vast majority of people to move to a fully flash based system.

      --
      Platform advocacy is like choosing a favorite severely developmentally disabled child.
    3. Re:Flash is ready even now by Leroy_Brown242 · · Score: 1

      I am no linux expert. Just a geek, who spends most his time with windows. (oh boy, I'd better put on my flame suit) But, with such a small distro, what can you run? With so little space, what users who's not a geek, will find a system usefull? I can't imagine surviving with anything less than 20GB or so these days.

    4. Re:Flash is ready even now by counterfriction · · Score: 1

      DS Linux is a perfect example of another linux distro that runs on flash alone.
      As for *limited* rights, its something like >5x10^6 writes to the card isn't it? I don't really know how that compares to the amount of r/w's a typical computer performs over a given amount of time, but it sounds like a lot to me. Also, with prices on flash memory falling as they are, it likely won't be a big deal to replace the card once you've expended your 5 mil writes.

      --
      Sig free's the way to be.
    5. Re:Flash is ready even now by Anonymous Coward · · Score: 0

      Thin clients, networked MP3/video/DVD playing HTPCs with networked storage, etc.

      If you have storage elsewhere on the network, it's not such an issue. Also if it's just a kiosk type machine for web/email BS it hardly needs massive storage either.

    6. Re:Flash is ready even now by ooze · · Score: 1

      I can't imagine surviving with anything less than 20GB or so these days.

      Ah, those were the days, when we had 10Mb harddrives. And a typical console Game took a 200k ROM.
      Yes, you can run _anything_ on 1Gb. Just remember, the 9Mb Emacs was once considered the epitome of bloatware. And that is a program you can actually do everything with.

      --
      Just because I can imagine doing a hippopotamus, doesn't mean I'd like to do it.
    7. Re:Flash is ready even now by bazorg · · Score: 1
      I like the idea of having a mini portable PC with the OS and most aplications packed up in a flash drive, then running everything from a RAMdrive.

      On the rare occasions you'd have to reboot, the flashdrive would be "read" for a big ISO file, and everything could run from, say, 1Gb RAM... that should be enough to hold and run Damnsmalllinux + openoffice.org + firefox, right?

    8. Re:Flash is ready even now by somersault · · Score: 1

      in the system you mention above, the NAND device sounds a bit pointless, if you only use it on first boot, and use another drive for long-term data storage. Also if your system crashes then you lose all your data - maybe you regularly shut down your PC but a lot of people dont. Having an 'original' file on the NAND drive is kind of pointless if it's a month out of date!

      --
      which is totally what she said
    9. Re:Flash is ready even now by somersault · · Score: 1

      look at http://www.damnsmalllinux.org/ - everything you need to be 'useful' in 50mb. It is because you use windows that you expect everything has to be inordinately wasteful - everything used to come on floppy, OS installs, application installs.. my first HD was a 180mb one for my A1200 . Why does everyone think that 'these days' everything should take up more space? Obviously if you want to store full sized pictures from your camera/mp3s/movies you need gigabytes of storage, but as far as an OS distro goes, you dont *need* a lot of space.

      --
      which is totally what she said
    10. Re:Flash is ready even now by Anonymous Coward · · Score: 0

      Just download Damn Small Linux or Puppy Linux and give it a try. It is amazing how much they fit in those distros compared to a normal linux install with similar functionality. Damn Small Linux is even set up so you can easily run it in an emulated environment under Windows.

      Personally I wouldn't use either of these as my main OS (they are too stripped down), but they are still impressive.

    11. Re:Flash is ready even now by m50d · · Score: 1

      http://cyti.latgola.lv/ruuni/index_en.html is a much better alternative to DSL. Full version of abiword rather than the diddly rtf editor you get with DSL. Full spreadsheet (gnumeric). Full copy of the gimp, mplayer...I could go on.

      --
      I am trolling
    12. Re:Flash is ready even now by somersault · · Score: 1

      hey I dont actually use DSL, was just pointing out how nobody needs to have a 2GB OS..

      --
      which is totally what she said
    13. Re:Flash is ready even now by yakovlev · · Score: 1

      Agreed. The correct way to handle this is:

      1.) Disable swap. Swap on a Flash drive is BAD.
      2.) Set automatic sync time to once per day (or hour, something "very long,") not the 30 seconds that it is in most machines.
      3.) Tune the filesystem victim selection to *strongly* favor pages that do not need to be written to disk.
      4.) Journaling BAD, Soft Updates GOOD.
      5.) In some environments, it may make sense to tune the filesystem such that, if one dirty block needs to be written to disk, write all dirty blocks to disk.

      The result of the above is a system that is more likely to lose data in a crash, but that does not totally break things, and which should at least preserve filesystem consistency. However, in return it minimizes flash writes.

      You can also play with making sync a no-op, but this usually isn't a good idea.

    14. Re:Flash is ready even now by m50d · · Score: 1

      Austrumi makes your point better though. DSL is clearly very stripped down, while austrumi is pretty much everything you want.

      --
      I am trolling
  9. Congratulations! by dbucowboy · · Score: 5, Funny

    Good job Samsung. I've got to NAND it to you...

    --
    This just in! 3 out of 4 people make up 75% of the population.
    1. Re:Congratulations! by tduff · · Score: 1
      Good job Samsung. I've got to NAND it to you...
      I think there should be a new moderation category: "Groan".
    2. Re:Congratulations! by Anonymous Coward · · Score: 0

      Would it be positive or negative, though?

    3. Re:Congratulations! by Anonymous Coward · · Score: 0

      And 3 out of 4 people are smarter than the dumbest 25% of the population.

    4. Re:Congratulations! by dbucowboy · · Score: 1

      haha... sorry. :p

      --
      This just in! 3 out of 4 people make up 75% of the population.
    5. Re:Congratulations! by zippthorne · · Score: 1

      It's not their fault really. They probably just have one of the 50% of all doctors who graduated in the bottom half of their class.

      --
      Can you be Even More Awesome?!
  10. Flash is a complementary technology, not a rival by BadAnalogyGuy · · Score: 5, Interesting

    Hard disks may be physically larger and slower for random access, but they are faster than Flash for large sequential reads, much in the same way that the hare is faster than the turtle in that old fable.

    We'll most likely see Flash storage grow in cell phones and PDAs, not in notebook computers. If you were a pilot, you wouldn't just have the mechanic swap out the propeller for a Rolls Royce jet engine. You'd want the whole plane overhauled to handle the increased stress on it. Better to have a system designed from the ground up that could handle the new engine rather than try to bolt it onto an older, proven design.

  11. Not a total replacement by Shamashmuddamiq · · Score: 4, Insightful
    Hard Drives will be useful for the forseeable future in lots of areas. Hopefully, however, in many applications, we can get rid of them altogether. With the correct wear-leveling algorithms, flash can last a long time. And there aren't big seek penalties like in hard drives, so read performance can be much better. And for applications where seek times dominate, this will boost performance big time. You'll be able to get good performance out of a fully normalized database without requiring nearly as much cache.

    I, for one, welcome... oh never mind.

    As flash drives become more and more popular, more dollars will pour into flash research and development. And applications will learn to accomodate the strengths and weaknesses of flash. I think we'll be seeing some really neat things over the next 10 years. Terabyte flash drive, anyone?

    --
    ...just my 2 gil.
    1. Re:Not a total replacement by johnnyoffline · · Score: 1

      What sort of implications would a hybrid storage situation pose? OS resides on flash and media/user files on hard disk? Does anyone have any insights on such an idea?

    2. Re:Not a total replacement by patio11 · · Score: 1

      I'm not a hardware guy, but it strikes me that this would be a nice thing to have in the various boxes that are sprouting up all over the place. I had a professor once that had some insane number of obsolete computers running Linux and some other OSS, performing all manner of Internet appliance type chores (firewall, intrusion detection, cycle server for student projects, what have you). Some of these seem like good candidates to yank out the hard drive from -- all the firewall needs is a Linux distribution and some configuration data to run, no need to put in a hard drive just so it can boot up and then come crashing down once in a while (the hard drive was essentially the only part he had to replace over the years, considering with 24/7 uptime it gets a bit of a workout and since the box had very, very little in terms of other components there were only a couple of places for it to fail).

    3. Re:Not a total replacement by thePowerOfGrayskull · · Score: 1

      Hmm.. I think you meant terabit.

  12. Re:I'm not surprised by iakie · · Score: 1

    hmm... I have to see the data sheet of the new chip before backing you. Right now I'm developing embedded system that uses small flash (2M-bytes) as storage. Apart from the bulk erase thing kept me headache (you can NOT erase/re-write data byte-by-byte, instead you have do that with a huge block of storage), the main problem with flash is its erase cycle. That is, you can only write/erase certain number of times before the block go dead. Sure, all storage have such problem, But with flash, the number is quite low. For smaller flash, it's usually 100,000 Cycle for each block. Keep in mind that Flash is a type of EEPROM, ie., it's a ROM. They're not meant to be written often.

  13. Re:Well ... by pdbogen · · Score: 1

    Because it will be half the (marginal, for the manufacturer) cost.

  14. Re:I'm not surprised by Sen.NullProcPntr · · Score: 1
    For nand flash it is on the order of 100k write/erase cycles. The file system is responsible for 'wear leveling' i.e. you write and erase a file 10 times the file will be written to a different part of memory each time.

    Most flash cards (SmartMemory etc..) and USB thumb drives are nand based.

  15. Re:Flash is a complementary technology, not a riva by joto · · Score: 4, Interesting
    We'll most likely see Flash storage grow in cell phones and PDAs, not in notebook computers. If you were a pilot, you wouldn't just have the mechanic swap out the propeller for a Rolls Royce jet engine. You'd want the whole plane overhauled to handle the increased stress on it. Better to have a system designed from the ground up that could handle the new engine rather than try to bolt it onto an older, proven design.

    It's not like it's something new and completely unproven. Solid-state disks (SSDs) have been used for years in server-applications, especially for large databases, where the speed of harddisks or RAID just won't cut it. This is an expensive solution, but if you have gazillions of transactions (think mastercard), it might still be cheaper than more traditional solutions (add more servers, add more disk-cache, make sure things don't fail).

    Given that it has worked pretty well at both the server-side as well as in gadgets and appliances, I'd say flash-memory notebooks are going to happen pretty soon. It's just a matter of hitting the right pricepoint. Today you can (theoretically) get a 2GB SSD for the same price as a 200GB HD. This is pretty uncool, although I would believe many enthusiasts would buy it, if there were producers of cheap SSDs (today only high-end SSDs exist).

    But if you could get a 20GB SSD for the same price as 200GB HD (which is a sane estimate, given the article), things start to make sense. It would be enough for running MS office on a laptop, and seriously reduce startup-time, as well as battery usage. Given it's performance, it would also be a great add-on for desktop computers (put the OS, most used applications, and swap-space on it, and use traditional harddisks for your videos/music/porn).

  16. Rival? by Transcendent · · Score: 3, Interesting

    How is a 2GB flash drive with only 100,000 erase cycles supposed to rival a much faster 500GB hard drive with a much, much longer life span? I think someone just wants to push their product...

    1. Re:Rival? by pilkul · · Score: 3, Insightful

      Much longer life span in principle, but if you get a lemon it might crash 2 weeks after you buy it... At least the flash memory will be able to warn you before it is close to expiry.

    2. Re:Rival? by Dun+Malg · · Score: 2, Insightful
      How is a 2GB flash drive with only 100,000 erase cycles supposed to rival a much faster 500GB hard drive with a much, much longer life span? I think someone just wants to push their product...

      Well, NAND flash like this is good for 1,000,000 writes rather than the 100,000 of NOR flash; but yeah, even that doesn't sound like enough. I don't know though. How much is enough?

      --
      If a job's not worth doing, it's not worth doing right.
    3. Re:Rival? by NanoGator · · Score: 1

      "How is a 2GB flash drive with only 100,000 erase cycles supposed to rival a much faster 500GB hard drive with a much, much longer life span?"

      Smaller/thinner/longer lasting laptops. I know we're not there yet, but if I could spend $300 for a 16 GB (byte, not bit) solid state drive, I'd happily plop it into my TabletPC. The desktop can be the storage device.

      --
      "Derp de derp."
    4. Re:Rival? by ihavnoid · · Score: 1

      1. Capacity : There are many people who can live with 5GB rather than 500GB.
      2. Lifespan : Think how many times you would erase a part of the flash in a year.

      At least, I would buy one if it provides 20GB @ $300. Better performance, better reliability, lower power, all at the cost of an tenfold size decrease, but I don't need that must space on my laptop.
      (I would still need a high-capacity hard drive on my PC, though.)

    5. Re:Rival? by Anonymous Coward · · Score: 0

      I think you guys got it all wrong! Imagine taking about 8 of these babies (that's 16 GB), each costing around $30 (yes not now but soon), and arranging them in a RAID 0 configuration. How fast do you think that will go in a laptop? Couple that with some SDRAM cache and you have yourself an awesome SSD. Or take some readily available 512 MB at $30 cards and arrange them as a RAID. A 2 GB SSD at $120 sounds pretty cool to me and considering peak read/write speed could be at 100MB/sec it is a steal. My guess is that "Inexpenisve" might be coming back into "RAID". Wikipedia

    6. Re:Rival? by Anonymous Coward · · Score: 0

      /longer lasting laptops.

      I don't need my laptop to last longer. It will last 5 years or more. I need it to be faster. If it is not faster, I will buy a new one.

    7. Re:Rival? by AaronLawrence · · Score: 1

      Why do you say that?

      --
      For every expert, there is an equal and opposite expert. - Arthur C. Clarke
    8. Re:Rival? by Alioth · · Score: 1

      With 1 million writes, that's more than sufficient for space to store data and programs, but not good enough for swap space. An (expensive) solution to swap space is to just add more RAM to the machine and turn off swap.

    9. Re:Rival? by v1 · · Score: 1

      how does it "warn" you? I know of no way to monitor how many flash cells have been 'spared', like you can do with hard drives.

      --
      I work for the Department of Redundancy Department.
    10. Re:Rival? by Anonymous Coward · · Score: 0

      2. Lifespan : Think how many times you would erase a part of the flash in a year.

      Every time you modify a file... Sorry, it's not as fast as a HD either.

  17. Re:I'm not surprised by 42Penguins · · Score: 1

    2GB per chip.
    It IS possible for the notebook to have more than one of these chips in it...

  18. Re:Hell Yeah! by name773 · · Score: 0, Offtopic

    Slashdot: nearly a million monkeys, but still no Hamlet.

    yeah, that's because it's over the posting limit (i know because i just tried to paste it in)

  19. peace and quiet by SimonInOz · · Score: 4, Insightful

    No moving parts = no noise.
    No moving parts = tough.
    No activity when quiescent - no heat.

    I, for one, welcome our new NAND overlords

    --
    "Cats like plain crisps"
    1. Re:peace and quiet by Lumpy · · Score: 1

      and not new. I have a Solid State IDE hard drive here from 1995.

      Solid State hard drives are an old idea that have been very VERY common in embedded and industrial systems for almost eternity.

      M systems makes really affordable IDE solid state drives in the 256 meg - 1 gig capacity that are tiny, realy low power and perfect for embedded systems. My car computer runs one for it's boot drive.

      Start researching embedded linux and embedded computing and you will find that lots of things you are just hearing about has existed for decades.

      --
      Do not look at laser with remaining good eye.
  20. Sandisk v Samsung et al by Douglas+Simmons · · Score: 1

    I'd expect an announcement for new higher calibar bar-lifting hardware performance to come from the company that does this pretty much exclusively (unlike Samsung and all the others who dabble in flash memory). That notwithstanding, I'd also expect the bigger guys who are spread out into other products to buy out a threat slash potential asset like Sandisk. Does Sandisk no longer have top-notch proprietary R&D having been beaten to the punch? Or is this a very specific kind of advancement that it isn't so significant.

    1. Re:Sandisk v Samsung et al by Anonymous Coward · · Score: 0

      Sandisk actually licenses from Sandisk, so they are not really "vs." fighting each other

    2. Re:Sandisk v Samsung et al by Anonymous Coward · · Score: 0

      ass in high res

      Hey that's a nice site. I shall add it to my list of faves.

      PS, nice pussy.

  21. OFFS! This is stupid. by pair-a-noyd · · Score: 4, Interesting

    Newegg, Maxtor 300 GIGABYTE sata $125. Available NOW.
    Vapordeals, Mysterymem, 16GB(?) $90. Available ???

    What's the R/W speed of these things? What's the R/W burnout on these?
    How many writes will they take before they fail?

    Maxtor is claiming a 1 million MTBF / 5 year warranty on their 300gb drive.
    No way in hell flash or any other memory is every going to compete with that,
    not in price, performance, capacity or endurance.

    Hard drives are so big and so cheap now that they are cheaper than blank DVD media. You're better off to archive to big drives then store them in fireproof safes than ANY other backup method. I have harddrives from the 80's that STILL have data on them that I can STILL retrieve and use, right now and I've made no serious effort to be overly protective of the drives. In other words, they've been kicking around the house in boxes on the floor. And they are still good. 20+ years later.

    Flash memory may have an indefinite SHELF lifespan but you can only write to them X number of times before they fail and they are slow.

    Someone is trying to sell the neophytes a bill of goods.
    When Vista releases there is going to be a rush to sell more silly crap to people. More upgrades.. Oh boy..
    In the meantime, I'll make due with my current system and my Linux.
    And as hard drives continue to get bigger and faster and cheaper I'll just add em as I need em.

  22. 11 years to replace 3.5 inch drives by matt21811 · · Score: 4, Interesting

    I did a study which estimates that flash will surpass 3.5 inch IDEs in every price by 2017.
    Read about it here:
    http://www.mattscomputertrends.com/flashvsharddisk .html

    1. Re:11 years to replace 3.5 inch drives by McFadden · · Score: 1

      If I was the CEO of Western Digital or Seagate I'd be having seriously sleepness nights about the future of my business. Those guys need to pump their R&D money into solid state storage research pretty bloody quickly before Flash memory does to them what WYSIWYG Word Processors did to Tip-Ex (aka correction fluid).

    2. Re:11 years to replace 3.5 inch drives by TheClam · · Score: 1

      Nah. Corporations are meant to die, just like the persons they're considered to be. When the goods they sell aren't in demand anymore, what'll they do, try to break into the samsung/hynix et al. marketplace? Nah.

      I say again, NAH! ;)

    3. Re:11 years to replace 3.5 inch drives by trandism · · Score: 1

      Nice... pffftt.. the term IDE in 2017 will only be used on magazines titled "Retro hardware"

      --
      www.lemonodor.com A mostly Lisp weblog
    4. Re:11 years to replace 3.5 inch drives by ricky-road-flats · · Score: 1
      Hmm, just read your article, and interesting though it is, there was such a big glaring mistake in your calculations (the first one I glanced at) I can't agree with your conclusions based on what you said.

      You say 18 months ago the fastest Flash drive you found in an Anadandtech article was 9 MB/sec, and that nowadays 23 MB/sec write and 30 MB/sec read is available. You then conclude this is a 300% improvement in 18 months!

      23 is only 2.56x9, and 30 is only 3.33x9, and that means the improvement is only 156% on write speeds and 233% on read speeds - impressive, but not 300% by a long stretch.

      FWIW, of course longer term spinning disks will be replaced, but I think it will be gradual. More and more machines are being built today without them, things like iPod nanos, HTPC machines which boot on flash to access media elsewhere, etc. and this will continue, as cameras, PDAs, phones and MP3 players all lose the need for spinning disks.

      At some point soon there will be useful-capacity flash drives which will become cheap enough to be attractive. I personally disable swap on all machines I own, so as soon as solid-state drives get to 20-30 GB or so, they will be great for simple machines. It will be a *lot* longer before the large storage servers I run are replaced by solid state, even through I'd love to.

      20 or 30 GB Flash at a reasonable price will happen within 3-4 years or so I imagine, and at that point laptops, office desktops and so on can use them. But replacing dozens of 146 GB and 300 GB uber-reliable SCSI drives? It will take a long, long time. Power and noise of hard drives don't matter in servers, as even very high-end drives burn 12 watts (I just had a quick look at a 146 GB 15,000 rpm one as a high-end example), and that is insignificant compared to 150W+ processors, dozens of fans, networking kit, KVMs, redundant PSUs, UPSes and so on.

    5. Re:11 years to replace 3.5 inch drives by olau · · Score: 1

      Interesting, but isn't extrapolating 11 years just a bit too long time to be meaningful? 11 years is enough time for another technology to pop up and be mass-marketed. Even if this does not happen, do you have any idea of the uncertainty of the estimate?

    6. Re:11 years to replace 3.5 inch drives by vertinox · · Score: 1

      Interesting, but isn't extrapolating 11 years just a bit too long time to be meaningful? 11 years is enough time for another technology to pop up and be mass-marketed.

      True, but wouldn't that mean Hard Drives would also be put out of the picture.

      My estimate is that we will have NAND OS boot drives for Vista and OS X by the end of 2007. Which will most likely push the whole flash drive for personal storage a few years early than his predictions.

      --
      "I am the king of the Romans, and am superior to rules of grammar!"
      -Sigismund, Holy Roman Emperor (1368-1437)
    7. Re:11 years to replace 3.5 inch drives by Surt · · Score: 1

      Wow, and your trends only have to hold true for another 11 years for the prediction to come true. I guess in an industry with as little innovation as computer technology, that just might happen!

      In all seriousness, the nand people don't have a good plan reaching out that far. In particular, by that point to maintain 100% avg capacity improvement every 2 years, let's call that 64x capacity improvement (I know, that would actually reach into the 12th year, but the graphs you claim are using improvement rates even faster than that: 160% per 2 years). That's an 8x improvement in process, and this memory was built in their new 50 nm process:

      http://www.samsung.com/us/Products/Semiconductor/U SNews/Flash/Flash_20050912_0000191464.asp

      That puts this new hypothetical future memory on their 6.25 nm process (or with your less generous demands, a 2.7 nm process). That's getting into the range of significantly less than 1000 atoms per device (evidence for atoms / device claim: http://gtresearchnews.gatech.edu/newsrelease/SiNW. html). To think there won't be any significant changes or difficulties in reaching that goal is optimistic.

      Meanwhile, all of this assumes that the hard disk people do nothing surprising in 11 years.

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    8. Re:11 years to replace 3.5 inch drives by matt21811 · · Score: 1

      I realise that you were making a joke but trends holding true for decades is not the unusual for the computer industry. If that wasnt the case then I wouldn't have bothered with all the research. That said, new developments could vary the timeframe dramatically either way or even render flash obsolete.

      Thanks for the additional information on process size, I'll use to guide my thoughts when I make further comments on my site. I will point out that shrinking the process may not be the only way to improve $ per meg figure for flash. The 10,000 fold increase in demand (thats just an guess of how extra megs need to be produced to replace hard disks) may help drive costs down signigicantly. The simple repetive layout of flash may also lead to improvements that are not avalable other types of silicon based equipement.

      "Meanwhile, all of this assumes that the hard disk people do nothing surprising in 11 years."
      I think a suprising thing is happening right now. Something that could make my prediction look conservative. Kryder's law seems to be broken. In a big way. Time will tell.

    9. Re:11 years to replace 3.5 inch drives by matt21811 · · Score: 1

      Thanks for the correction. I'll be fixing that up shortly. The change doesnt alter the fact that if this rate of improvement continues then flash will actually pass disk in performance way before 11 years. Which is what I was trying to show.

      You are entirely correct that flash will eat away the disk market starting at the areas where disk offers the worst megs per $.

      The reason I chose 3.5" IDE drives for my survey was that they offer the very best $ per meg you can get. If flash surpases these disks then all disks are pretty much gone. I was really trying to find out roughly when we would be saying bye to disk drives.

    10. Re:11 years to replace 3.5 inch drives by Surt · · Score: 1

      Yeah, the main point of interest is really process size ... silicon devices are rapidly headed for a dead-end in terms of device size ... conventional electronics cannot possibly operate any smaller than one atom, and 11 years from now we'll not be far removed from that (and frankly, it is difficult to imagine how even a 10 or 100 atom device would operate reliably). So I'd definitely expect progress to level off in the coming years as we approach that limit. Either we'll wind up switching to some completely non-conventional type of device (spintronics, qubits, etc, in which case I think you'd probably have to agree all bets are off) or we'll really be stuck. Interestingly enough, projecting as little as 30 years into the future: either we've got subatomic electronics (something of a misnomer there as electronics applies atomic level) or we've stopped being able to continue doubling density even at an every-3-years slower-than-current doubling rate.

      Personally, I suspect that computer technology is headed for a wall.

      On the other hand, if we get there, we'll have Trillion transistor chips. Hopefully 640 Terabytes of RAM is enough for anybody.

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    11. Re:11 years to replace 3.5 inch drives by Anonymous Coward · · Score: 0

      So what I did a study that proved the space program was a waste of time since by 2017 trees will grow to the sky.

  23. Re:Flash is a complementary technology, not a riva by Anonymous Coward · · Score: 2, Informative

    I think it's worth mentioning that the bottleneck in reading/writing large files is an interface problem (usb et.al.) and not actually an issue with the ram. Currently the thing spinning drives have going for them is cost per GB.

  24. Apples ,Oranges,Chips,Disks, and Drives. by mysterystevenson · · Score: 2, Insightful

    Remember your first Nintendo "Mario's"? Plug em in n out, not enough gold on the contacts, but the chips still play if you blow on the contacts.How many disk games have worn out since? How many Hard drives still work with the old games on em? Maybe optical storage or quantum computers will come along., but I can see what has lasted the longest so far.

    --
    MYSTERY
    1. Re:Apples ,Oranges,Chips,Disks, and Drives. by NanoGator · · Score: 1

      'Remember your first Nintendo "Mario's"? Plug em in n out, not enough gold on the contacts, but the chips still play if you blow on the contacts.How many disk games have worn out since?"

      Amusingly, the reason why blowing on the contacts worked was because you were causing condensation to accumulate on the contacts. The reason the contact would lose connection is that if you rub two sheets of copper together, it produces a buildup of this dark gunky substance. Instert a cartridge a few times and enough buildup happens to break the connection. Clean the cartridge (or make the contacts wet, as you suggested) and the connections will light up again. This problem went away with every Nintendo catridge system ever since.

      Anyway, here's the part where I agrue with you: EVERY medium you described is read only. Sure that'll last for decades, but that's not suitable for general computing purposes. With today's technology, reliability is still a huge issue. That's not to say that they won't become super resilient in the years to come, but you do have to consider that you're talking about technology with constant state changes. Those will eventually break. In simpler terms, the light switch in your bathroom will have a much shorter life than the wires carrying the electricity to the bulb.

      That said, I can easily imagine these things eventually overtaking HDs in reliability. So it's not like we totally disagree. :)

      --
      "Derp de derp."
    2. Re:Apples ,Oranges,Chips,Disks, and Drives. by ShyGuy91284 · · Score: 1

      Sounds about right. A friend of mine (I never had a NES, all my friends did, and i wanted one... *cry*) actually I think exhaled on to them (like one might to fog up glasses and clean them) instead of blow on them. Seemed to work.

      --
      In undeveloped countries, the consumer controls the market. In capitalist America, the market controls you.
    3. Re:Apples ,Oranges,Chips,Disks, and Drives. by theLOUDroom · · Score: 1

      Remember your first Nintendo "Mario's"? Plug em in n out, not enough gold on the contacts, but the chips still play if you blow on the contacts.How many disk games have worn out since?

      Actually it's usually the connector that wears out. It you replace it, you probably won't have to blow anymore.

      --
      Life is too short to proofread.
    4. Re:Apples ,Oranges,Chips,Disks, and Drives. by poot_rootbeer · · Score: 1


      Nintendo cartridges stored their code in ROM, which degrades very very very little with usage and time, and has been that way since it was invented 35 years ago.

      It's not really comparable to Flash memory in any way.

  25. Re:Flash is a complementary technology, not a riva by hyc · · Score: 1

    You're comparing apples and oranges. The solid state disks you're referring to all use regular DRAM with battery backup. DRAM is a few thousand times faster than flash.

    Flash-based drives aren't even up to UDMA66 speed yet. For notebooks, my 60G Hitachi 7200rpm drive will be faster than flash in every situation.

    --
    -- *My* journal is more interesting than *yours*...
  26. Re:OFFS! This is stupid. by wall0159 · · Score: 1

    But you're comparing a desktop hard drive solution to one which is designed to be ultra-portable. How much does a 300GB 2.5" drive cost? (hint: current max size is about 120GB)

    For a lot of portable applications, 20 - 30GB of storage is plenty. Esp if it's the size of a credit card and uses 1/10 the energy.

    I'm sure speed issues can be circumvented be having some kind of striped (RAID3, is it?) configuration of the individual chips...

  27. 2GB SD cards already out by Neo-Rio-101 · · Score: 1

    If 16Gb = approx 2GB, and there are already 2GB SD Flash memory cards available, does this mean that we could end up with 4GB -> 8GB SD memory cards in the forseeable future?

    (Eagerly awaiting next iteration of the Sharp Zaurus!)

    --
    READY.
    PRINT ""+-0
    1. Re:2GB SD cards already out by v1 · · Score: 1

      you can already get 4GB SD, compactflash, and usb thumb drives.

      --
      I work for the Department of Redundancy Department.
  28. 16 gigabit? by Teque5 · · Score: 1, Redundant

    is it just me or is 16gigabits = 2 gigabytes? 8bits = 1byte AND - dont they already have 2gig flash chips? teque5.95mb.com

    --
    teque5.com
  29. Not Exactly True by Jack+Kolesar · · Score: 1

    If you are going to be a stickler, then you should realize that "B" used to stand for and sometimes still stands for bit. Granted it is much more common nowadays to mean byte. But, that wasn't always the case. If you are going to get technical, it may be better to use OCTET.

    http://www.answers.com/topic/byte

    1. Re:Not Exactly True by Anonymous Coward · · Score: 0

      Uh? "B" used to stand for and sometimes still stands for *bel*.

      http://www.answers.com/topic/bel

    2. Re:Not Exactly True by AcidPenguin9873 · · Score: 1
      Please stop posting this crap to Slashdot.

      http://en.wikipedia.org/wiki/Byte#Abbreviation

  30. Re:OFFS! This is stupid. by Anonymous Coward · · Score: 0

    Endurance? Hahaha, don't make me laugh. Hard drives are NOT known for their endurance. I've had 5 hard drives die within the past 7 years.

    Let's see.. one Hitachi notebook drive died when I accidentally bumped the notebook pretty hard against the table. Then there's the one that died right after a power outage (Purchased a UPS right after that). Also there's the ones that just.. die. Of course, that was a Maxtor, which has a notoriously bad reputation. The list goes on.

    Flash based hard drives will be a huge improvement in terms of durability. Notebooks really need this right now, more than anything. Screw huge storage space, none of that matters if your hard drive suddenly dies on you.

  31. Re:Flash is a complementary technology, not a riva by markdavis · · Score: 1

    The solid-state drives you mention are typically based on battery-backed up, volatile, RAM. Sometimes coupled to a hard drive... something akin to a REALLY huge hard drive cache. They are not based on Flash. Flash memory has a limited write cycle, and is *MUCH* slower in reads than fast SCSI RAID storage.

  32. Re:I'm not surprised by pilkul · · Score: 1

    Hard drives are still about 300x bigger for the same price. Flash is only going to edge forward in lightweight portable applications.

  33. Flash does get firkled by bdwoolman · · Score: 2, Insightful

    As much as I prefer flash for an MP3 player over a HDD I admit that funny stuff happens to it. I have had some camera CF cards do peculiar things. Especially if there is a power problem when they are writing. Remember the mars rover was hamstrung for awhile with a flash problem. They sorted it though.

    --
    "No fear. No envy. No meanness." Liam Clancy
    1. Re:Flash does get firkled by Slashcrap · · Score: 1

      As much as I prefer flash for an MP3 player over a HDD I admit that funny stuff happens to it.

      I think I'm going to stick with hard drives then, as I've never had one of them do anything funny. I can't even imagine a technology that would be more reliable than good old hard drives.

      I have had some camera CF cards do peculiar things. Especially if there is a power problem when they are writing.

      Which is obviously an OS / filesystem issue.

      Remember the mars rover was hamstrung for awhile with a flash problem.

      I'm pretty sure that was an OS issue too. It just manifested in the flash getting filled up.

    2. Re:Flash does get firkled by Rothron+the+Wise · · Score: 1

      Remember the mars rover was hamstrung for awhile with a flash problem.

      It's true that the mars rovers uses flash, but the problem one of the rovers suffered can hardly be called
      a "flash problem". It was a file system problem. The FAT they used on the flash had a limit to how many files it
      can hold in one single directory. http://marsrovers.jpl.nasa.gov/newsroom/pressrelea ses/20040201a.html.

      --
      A witty .sig proves nothing
    3. Re:Flash does get firkled by Anonymous Coward · · Score: 0

      What funny things happen? I know you have to be careful and make sure they are properly unmounted before you disconnect them or they'll get corrupted. I just got one for Christmas and haven't had a problem with it yet.

  34. Read Flash Specs by putko · · Score: 1

    If you read the specs on typical flash memory, the stuff can last around 10 years until the data isn't necessarily correct.

    Doesn't that make them about as good as the dye based CD-Rs that people fear will not be good for archives?

    I'm not saying I'm the expert on this -- I'd appreciate it if someone could explain to me that the flash memory will actually last 40 years or so. But I doubt it.

    --
    http://www.thebricktestament.com/the_law/when_to_s tone_your_children/dt21_18a.html
    1. Re:Read Flash Specs by trandism · · Score: 1

      40 Years??!!!!111111

      What the fuck for??
      In 40 years time we 'll be able to store stuff in god knows what?? I won't be surprised if the standard storage will be dust on the corner of your mama's basement, USB-35 mobile condomns or the hair of our assholes..

      --
      www.lemonodor.com A mostly Lisp weblog
    2. Re:Read Flash Specs by putko · · Score: 1

      Well, if your firmware is in flash, it matters.

      I wouldn't want a car to stop working in a few years just because the flash went bad.

      --
      http://www.thebricktestament.com/the_law/when_to_s tone_your_children/dt21_18a.html
    3. Re:Read Flash Specs by trandism · · Score: 1

      Sure pal but grand-grand-grand-parent said 40.. not a few... you keep cars for 40 years? well in that case...

      --
      www.lemonodor.com A mostly Lisp weblog
    4. Re:Read Flash Specs by putko · · Score: 1

      The Specs say 10. If you are lucky it might be 40 years.

      But if it is only ten, that could easily be the part that breaks your car.

      It is worse than this: most microcontrollers are flash-based. E.g. the Atmel ones, which are the most popular.

      The life is 10 years, which is too short for their product lifetimes.

      When microcontrollers fail, people can get killed.

      --
      http://www.thebricktestament.com/the_law/when_to_s tone_your_children/dt21_18a.html
  35. Secured OS? by Anonymous Coward · · Score: 3, Interesting

    When I see some more knowledgable people here talking about putting the OS on flash, reading it to active memory, etc. the first thing that comes to mind (especially given the limited writes) is that it's the right technology to introduce the "trusted" OS - you don't want to write to the OS unless it's a patch, you don't let the user change much if anything, it's in a stable form that is quickly loadable, probably faster that OS on harddrive now that is the technologic foot in the door to entice people to upgrade to it. If they limit the writability with some sort or authorization scheme that is changable with the update, ANY writing can be limited to over a connection to the home server. Sure, a business wouldn't like/allow/purchase that, but I think that the whole trusted computing idea is aimed at controlling home computers, not the business machines, yes? It would seem to be an elegant fit with (my understanding of) trusted computing.

    1. Re:Secured OS? by marcosdumay · · Score: 1

      Nothing that can't be done with disks. Get it easy, we are just talking about replacing a disk with a chip. Anything can go on a chip, as anything can go at the disk's controller.

  36. Re:OFFS! This is stupid. by pilkul · · Score: 1
    Of course, that was a Maxtor, which has a notoriously bad reputation.

    Meh, every hard drive manufacturer has a "notoriously bad reputation" to somebody. Specific lines of hard drives can be lemons if there's some small defect in the production process, and no manufacturer has always been immune to this. IBM had its DeathStar, etc.

  37. Re:Flash is a complementary technology, not a riva by benzapp · · Score: 1

    No, I think you are quite mistaken. Read up on Hybrid Drives.

    The future of laptop hardrive technology is going to be a mixture of hard disk and flash memory technology.

    --
    I don't read or respond to AC posts
  38. Re:OFFS! This is stupid. by pair-a-noyd · · Score: 1

    It's regretable that you've had such bad luck with hard drives.
    It happens to some people.

    I on the otherhand, have had extremely good luck with hard drives,
    I've had a few die on me but they were Western Digital. WD is CRAP. I've
    gone through a LOT of WD drives and I finally learned my lesson, I'll NEVER buy another WD, EVER..
    I fried a few by putting the power cable on backwards with the PC running, but that was my dumb fault.
    As for Maxtor, people say they are crap but I've had no complaint with them at all.
    I guess I'm just luckier than most..

  39. Re:OFFS! This is stupid. by markdavis · · Score: 2, Insightful

    >Newegg, Maxtor 300 GIGABYTE sata $125. Available NOW.
    >Vapordeals, Mysterymem, 16GB(?) $90. Available ???

    No, it is 16Gb (2GB), not 16GB. That is an order of magnitude larger... making your argument even more correct. Flash memory will probably never (in the foreseeable future) overtake or even come close to what you can do with rotating magnetic media.

    Someday there might just be a 20GB flash drive for $200, and at that point in time, there will probably be a 20TB hard drive that is 10 times faster for $200, or a 2TB hard drive that is 10 times smaller and uses 10 times less power than the 20TB for $200.

    Still, seeing higher densities of flash is a wonderful thing, especially for portable/pocketable devices. I am AMAZED when I see how inexpensive a 1GB tiny SD card is! It would just blow my mind away 10 years ago... Bring it on!

  40. Re:OFFS! This is stupid. by matt21811 · · Score: 4, Insightful

    Nice rant but you totally mised the point.
    A 300 Gig IDE drive doesnt fit in a laptop.
    A 300 Gig IDE drive uses loads of power.
    A 300 Gig IDE drive has faster sustained transfer speed but much a longer access times than flash. Horses for courses.
    Wear leveling algorithms can make the write limit of flash irrelevant.
    That the interface (eg, ATA) for accessing storage media usually goes out of date before the media wears out is true for both disks and flash.
    The real story here is that flash is trouncing disk in improvements in Megs per $ and will one day catch up to and overtake disk. And it will be sooner than mmost people expect.

  41. Gigabit? by homeobocks · · Score: 0

    Fuck the decietful asswipes who use gigabits for measuring storage and call 10^9 bytes a gigabyte (instead of 2^30). 16 Gb sounds like something until you realize that it's 2 GB.

    --
    MOUNT TAPE U1439 ON B3, NO RING
    1. Re:Gigabit? by be-fan · · Score: 1

      They use gigabits, because like RAM modules, a flash module contains multiple memory chips. Your 256MB RAM stick is usually made using 8 256 megabit modules. Similarly, the 16 gigabit flash chip will be used in devices featuring a lot more than 2GB of total storage.

      --
      A deep unwavering belief is a sure sign you're missing something...
    2. Re:Gigabit? by 91degrees · · Score: 1

      By definition, the "Giga" prefix means 10^9. Just because RAM manufacturers have been approximating for so long doesn't make them right.

    3. Re:Gigabit? by raygundan · · Score: 1

      A-freaking-men. We have a new standard to replace our power-of-two bastardization of the SI prefixes for powers of 10. Just because we've been doing it forever doesn't mean the computer geeks are in the right-- we clearly took an established international standard for the meaning of prefixes and changed it. As dumb as the new prefixes sound to my ears, it's time we stepped up and got used to it.

      You'd think that programmers, of all people, would value the clarity of meaning something like this would bring-- but it just goes to show that even those of us who are logical for a living have things we're irrationally attached to.

  42. Re:Hell Yeah! by ihavnoid · · Score: 2, Insightful

    I'd rather add that much RAM on my PC, since they are faster with a similar price.

    You definitely should consider that RAM price is dropping as fast as flash memory price.

  43. Not replace due to limited # of writes... by Zantetsuken · · Score: 1

    since flash memory has only a limited number of drive writes, there isnt a chance in hell it will replace conventional hard drives - rival conventional drives in archival purposes, maybe - but with only a limited number of writes, what do you do when it goes bad because you wrote to it to store the database for your forum?

    1. Re:Not replace due to limited # of writes... by pe1chl · · Score: 1

      Flash disks solve this problem by using a translation table between the blocknumbers that you write to, and the real flashblocks on the chip.
      When you write to a block many times, another block is picked on the chip and the blocknumber is written to the table (also in flash). Then you start wearing out a different block on the chip.
      With a clever manager routine this can balance the number of writes all over the chip, and it looks like you can write billions of times to one single sector.

  44. Fear my ... by l33tlamer · · Score: 1, Informative

    2.4 Terabit Hard Drive, or in layman non-marketing speak, 300GB... I mean 279.4 Gigabytes.

    I would like to patent a new measurement unit, called the bi (pronounced "bee"). Not to be confused with the slang used to describe a person with bisexual tendencies, this measurement unit quantifies memory size.

    1 bi = 0.001 bit

    Fear my 2400,000,000,000,000 bi hard drive!!!111

    --
    If I can do it, its probably not worth doing... probably
    1. Re:Fear my ... by Anonymous Coward · · Score: 0

      You don't wanna know how many micronibbles fit in my disks

  45. Oh no by Anonymous Coward · · Score: 0

    is it just me or is 16gigabits = 2 gigabytes?

    No, it's just you.

  46. write-only by commodoresloat · · Score: 1
    There is only so many times (in the 10s of thousands)

    Yes but flash will overtake hard drives in most Write-Only Memory applications.

  47. Offtopic by nexcomlink · · Score: 1

    I know this is an off topic question but I remember hearing that the macbook pro uses some type of NAND flash to boot up instead of using the HD. Now if what people say that NAND flash dies after certain read/writes then that would mean a limited amount of restarts as well. Correct me if I am wrong though.

    1. Re:Offtopic by damned_mediocrity · · Score: 1
      From what I understand, that was merely a rumor. Intel had demo'd a new flash-based caching technology a while back, and all the Macheads jumped on this as a sign of things to come in Apple laptops. I doubt it materialized -- if the MacBook used NAND flash for booting up, wouldn't Steve-o have said something about it in his keynote?

      Who knows, though. The damn things don't ship until February in any case.

    2. Re:Offtopic by SorcererX · · Score: 1

      They handle unlimited reads, they don't handle unlimited writes/erases though, so after a few million Mac OS X updates, the NAND might have to be replaced :)

      --
      Any sufficiently advanced technology is indistinguishable from magic.
  48. Flash memory is too tempermental by themadplasterer · · Score: 1

    While flash memory may seem ideal it is terrible in dealing with power interuptions when the filesystem is mounted. Just ask apple how many ipod shuffle's and nano's were rendered useless by a power outage. We all know that power interuptions can wreak havoc with any electronic device, but it seems to me that you gotta wear kid gloves with flash

  49. Re:Flash is a complementary technology, not a riva by earnest+murderer · · Score: 1
    I think it's worth mentioning that the bottleneck in reading/writing large files is an interface problem (usb et.al.) and not actually an issue with the ram. Currently the thing spinning drives have going for them is cost per GB.

    100+MB read and about 50 MB write last I read.
    For comparison, a Seagate Barracuda 7200.9 3.5" desktop drive will do about ~47 MB per second.

    --
    Platform advocacy is like choosing a favorite severely developmentally disabled child.
  50. Re:Flash is a complementary technology, not a riva by Anonymous Coward · · Score: 2, Funny

    If you were a pilot, you wouldn't just have the mechanic swap out the propeller for a Rolls Royce jet engine.

    Well done, you have truly lived up to your username.

  51. Re:I'm not surprised by Anonymous Coward · · Score: 0

    The problem is that's what the article is talking about, it being used more, to replace HD's in portables.

  52. Larger disk space needs counterbalance this by justine_avalanche · · Score: 2, Insightful

    Sure the Flash memory are growing in size at some rate r per year, but at the same time the need for more and more disk space is also growing at some rate r'.
    I can't say if r > r' so much that in the course of the next few years we'll see HD disapear ... I doubt it.

  53. Number 1 reason why not to use flash chips by Anonymous Coward · · Score: 0


    They are susceptible to ESD. Data can be lost in an instant.

      CDs and IDE Hard Drives are more reliable.

  54. Explain to me... by ashpool7 · · Score: 1

    ...how "wear leveling" solves any problems. I get the concept: http://www.dataio.com/pdf/NAND/MSystems/TrueFFS_We ar_Leveling_Mechanism.pdf but maybe I'm reading this wrong.

    Considering your "virtual erase unit" (VEU) to be a multiple of the physical erase unit, you still have to have a significant number of these VEUs so you can remap things, lowering the flash density as you increase the reliability. Make the VEU too large, and the leveling algorithm doesn't give you that much time. Make it too small, the flash chip is now twice the size. Not only that, but in order for the VEU to remember where to map what, it itself has to be flash. What happens when the VEU flash area gets worn out? How would you even detect that the VEU is getting written too much without, oh, keeping track of THOSE times in flash too.

    And what happens when the disk is mostly full and there are only a few empty sectors to write to, as most data gets overwritten? Guess you're screwed then!

    No, I don't think this is a very good trade. Flash has a physical problem that needs to be solved with materials science, not computer science.

    1. Re:Explain to me... by vidarh · · Score: 1
      What you are assuming is a physical problem only is a physical problem if you use the device the wrong way. This is like saying that hard drives have a physical problem that needs to be solved because turning the drive motor on and off all the time makes it wear out quickly.

      I also don't think you understand how the wear levelling you referred to works.

      Generally most flash file systems are variation of journalling systems where you maintain (and regenerate on startup) an in-memory map mapping the newest version of each block to a specific physical block on the device. That in itself instantly reduces wear, because rewriting a block means writing a new physical block, not rewriting the same location.

      Then, if you are running low on space, an algorithm is used to pick one or more physical erase units to "garbage collect" - that is, you in a set of erase units, remove all blocks from them that have been remapped elsewhere on the device (because they have been rewritten later), erase the units and write back a compacted set of blocks, including meta data that will allow you to quickly determine which physical blocks are mapped where on device startup.

      So the density will typically remain more or less the same. The only problem with this system is if the device gets very full and writes only happen to a single area. But this can also be sorted out by also occasionally remapping blocks that are not actually being rewritten by the user, slightly increasing the overall number of erases for the device as a whole, but keeping the number of erases for a single block down.

      The overall result of this is that you are unlikely to ever run into the wear problem in systems that aren't SEVERELY IO bound. That is, your desktop/laptop is likely to die years before the flash drive, but you might want to avoid it for your highly loaded database server / mail server.

  55. Re:Hell Yeah! by ePhil_One · · Score: 1
    I'd rather add that much RAM on my PC, since they are faster with a similar price.

    1. Unless you are running a 64 bit system, you can only address 4GB
    2. When the power goes out, you'll need to reinstall your system

    --
    You are in a maze of twisted little posts, all alike.
  56. 10,000 writes/second for 13 years by morcheeba · · Score: 4, Informative

    That's 100k per block, not for the entire drive. The wear-leveling algorithms will make sure that even if you constantly re-write the same file, that part of the memory won't get worn out.

    With a 512-byte erase block size, that is 419 billion writes. With a 4K erase block size, that's 52 billion writes. Use a 20GB drive instead of 2GB, and you'll get 10x the writes. And, the computer can warn you before the memory stops re-writing.

    5 trillion writes is 10,000 writes/second for 13 years.

    1. Re:10,000 writes/second for 13 years by somersault · · Score: 1

      though surely the more data that you are storing long term, then the less space you will have for spreading around the writes, so if you are actually using most of your 20GB drive (which isnt very difficult to achieve), then you will wear out one section a lot faster. And then if you arent using most of your 20GB drive, why did you spend all that money on 20GB of NAND in the first place?

      --
      which is totally what she said
    2. Re:10,000 writes/second for 13 years by anum · · Score: 1

      To you that data is static, but your harddrive/flashdrive logic controler can move that data around all it wants. Fragmentation won't be a big deal on a flash drive so this logic can function at the individual sector level. When one sector has been static too long it will just move the data to a well used sector and free it for more dynamic data. Not as straight forward as "randomize new writes to free sectors" but not that complcated I suspect.

      --
      I don't think, Therefore I'm not.
    3. Re:10,000 writes/second for 13 years by tootlemonde · · Score: 2, Informative

      so if you are actually using most of your 20GB drive (which isnt very difficult to achieve), then you will wear out one section a lot faster.

      Wear-leveling algorithms can take this problem into account. According to this article on Solid State Disks (SSD):

      When a given block has been written above a certain percentage threshold, the SSD will (in the background, avoiding performance decreases) swap the data in that block with the data in a block that has exhibited a "read-only-like" characteristic.
    4. Re:10,000 writes/second for 13 years by somersault · · Score: 1

      I did realise that it could move the data if one area is getting rather overused, but I havent heard anyone say that it moves sections of unused data around to save on the device's life. If it does then that's fine and dandy

      --
      which is totally what she said
    5. Re:10,000 writes/second for 13 years by dmccarty · · Score: 1

      So you're saying I'm screwed in the 14th year?

      --
      Have fun: Join D.N.A. (National Dyslexics Association)
    6. Re:10,000 writes/second for 13 years by poot_rootbeer · · Score: 1

      5 trillion writes is 10,000 writes/second for 13 years.

      And good luck finding a Winchester drive that hasn't developed some bad sectors after 13 years of continuous service.

      The limited-rewrite-cycle of Flash memory is pretty much a red herring. It's never going to affect 95+% of users.

  57. Re:OFFS! This is stupid. by earnest+murderer · · Score: 1
    Flash memory may have an indefinite SHELF lifespan but you can only write to them X number of times before they fail and they are slow.

    Yes, because 1 million writes is so not worth the 100+ MB per second read/ 50 MB per second write speed. That's the industry standard. For some extra money you can bump that to 5 million with a 5 year warranty.

    What does my brand new laptop do? About 20. Even the newest fancy pants Raptors peak at 80 MB per second at the edge of the disk.

    Considering how late your comment is, you should have read all of the above by now. Even still, if you didn't know these things you should know better than to spout off like some kind of funduhmentalist christian. Do you really believe the scientists and business people with all of their learning and experience have missed out on an obvious truth that only you know about.

    Know what?

    Digg.com called, they want their "conversation" back.

    --
    Platform advocacy is like choosing a favorite severely developmentally disabled child.
  58. How bout NOR Flash? by yuriismaster · · Score: 1

    You know, I wonder, can someone please make NOR flash? I mean it's functionally complete too. Anyone know of any available NOR flash?

    I wonder if I'm misinterpreting the meaning of NAND in this instance, where the hardware uses chained NAND gates to store values. Can't the same design be implemented using NOR's?

    1. Re:How bout NOR Flash? by LOTHAR,+of+the+Hill · · Score: 2, Informative

      The traditional flash devices are NOR flash devices. They are by nature more reliable than NAND flash but are expensive to amnufacture. NOR flash devices can withstand more write write cycles than a NAND flash. BIOS flash are NOR, as are most of the solid state flash drives used in telecom and aerospace. This is why NOR flash is marketed to embedded/industrial customers and NAND flash is marketed to the consumer market.

    2. Re:How bout NOR Flash? by Anonymous Coward · · Score: 0

      Neither NAND nor NOR is better.

    3. Re:How bout NOR Flash? by Anonymous Coward · · Score: 0

      Nor NEITHER nand NOR is not

    4. Re:How bout NOR Flash? by hyc · · Score: 4, Informative

      NOR flash is extremely slow for writes. This Samsung appnote
      http://www.samsung.com/Products/Semiconductor/Memo ry/appnote/onenand_features_performance_051104.pdf
      compares I/O performance of the various technologies (the chart is on page 28, so scroll down...)

      For their test rig, NAND flash yields 8.8MB/sec writes vs NOR at 0.14MB/sec. That's why NOR flash is only used for BIOS memory and other things you don't have to rewrite very often. On the flip side, NAND flash gets reads at 16.5MB/sec vs NOR at 23.9MB/sec (or 108MB/sec, presumably in some kind of burst mode - that part isn't explained).

      If their OneNAND performs as well as they claim, I could see using it for a boot drive; 68MB/sec read would be fine there, 9.3MB/sec write would be ok as long as you weren't paging to it or doing much of anything else. Linux would run pretty well with those parameters, its buffer cache is good at absorbing and deferring writes; Windows 2K/XP's memory manager/cache manager purges pages too aggressively though, which would make the write throughput a serious system bottleneck.

      --
      -- *My* journal is more interesting than *yours*...
    5. Re:How bout NOR Flash? by Kuj0317 · · Score: 2, Informative

      the number of transistors required to make a nand gate is less than the number required to make a nor gate. Therefore, nor would offer no benefit, while requiring more tx's (higher power consumption and less chip density)

    6. Re:How bout NOR Flash? by Anonymous Coward · · Score: 0

      Linux would run pretty well with those parameters, its buffer cache is good at absorbing and deferring writes;

      Yes it is very good, which is annoying when I copy lots of MP3s to my MP3 player and then have to wait 5 minutes for it to flush the cache when I go to unmount it. I tried using mounting using the sync option and that made it continually stop and start, so the copy takes longer, at the moment my solution is to copy using the CLI and unmount the drive immediately after the copy is complete.

    7. Re:How bout NOR Flash? by Non+Creative+Guy · · Score: 1

      Speed of any logic gate is dependent on the longest string of series connected transistors. In NOR gates that string is made of pFETs and in NAND gates it is nFETs. Since nFETs are faster (higher mobility for electrons than holes) NAND gates are faster.

    8. Re:How bout NOR Flash? by hyc · · Score: 1

      That's why we have a "sync" command. On Unix/Linux, you have complete control over whether to favor cache or to favor disk consistency. On Windows, you have no choice, you're stuck with whatever policy Microsoft dictates to you. I think that accurately reflects the state of affairs at all levels.

      --
      -- *My* journal is more interesting than *yours*...
  59. Re:OFFS! This is stupid. by pair-a-noyd · · Score: 1

    Yes indeed on all accounts.

    As for what you can do right now, ONE TB will run you about $500.
    Just 5 years ago that amount of storage capacity was SciFi movie stuff of the future.
    Now it's only $500 and 3 days UPS ground away from reality.
    And with HTPC here, 1TB isn't all that big of a deal.
    SERIOUS HTPC people will probably want 2-3TB..

    As for notebooks/laptops, 2.5" drives are growing/shrinking too.
    I do not believe for a minute that NV memory drives will ever replace mag drives.
    I've seen these nonsense pie in the sky promises before.
    I have one of them right here.
    Read about it here --> http://hardware.slashdot.org/comments.pl?sid=17330 3&cid=14420399

  60. Re:OFFS! This is stupid. by Schraegstrichpunkt · · Score: 1
    WD is CRAP. I've gone through a LOT of WD drives and I finally learned my lesson, I'll NEVER buy another WD, EVER.. I fried a few by putting the power cable on backwards with the PC running, but that was my dumb fault.

    IDE drives aren't hot-swappable, backwards or not, and you need to take proper precautions against electrostatic discharge.

  61. this is crap. by Anonymous Coward · · Score: 0

    Sony has with 4 Gig pendrives. It has problems for now, ie cant transfer files >1gigbyte and only one file at a time. It was manufactured in China and offered to employees at approx 70USD. So this news is crap.

    1. Re:this is crap. by HaydnH · · Score: 1

      They will be made up of multiple chips smaller than those in tfa, probably 1Gb, so this news will apply to those as well.

      --
      Time is an illusion. Lunchtime doubly so. - Douglas Adams
  62. Erm... ? by renehollan · · Score: 1
    What "Erm..." ?

    Beats swapping over the 'net to NFS...

    --
    You could've hired me.
    1. Re:Erm... ? by Aardpig · · Score: 1

      Not when you realize that flash has a finite read/write count for each block.

      --
      Tubal-Cain smokes the white owl.
    2. Re:Erm... ? by renehollan · · Score: 1
      Not when you realize that flash has a finite read/write count for each block.

      Noted, but proper flash filesystems can address this by relocating rewritten blocks on the device.

      Though, it was amusing to think about a swap-to-flash web server getting a good ole' slashdotting, and burning up... literally.

      --
      You could've hired me.
  63. Re:OFFS! This is stupid. by Anonymous Coward · · Score: 0

    You seriously miss the point, and not just because you're comparing a desktop solution to a notebook solution.

    Flash will be the storage medium of the future. I'm working on several proposed implementations now. Yes, there are issues to overcome, but the number of writes is not one of them.

    Hard drives are big and cheap now, there's no doubt about that. But flash drives will become big and cheap at some point. The slowest flash memory is faster than the fastest hard drive.

    And your MTBF canard shows that you don't understand that MTBF is a measure of reliability, not endurance. That 1,000,000 hour MTBF means that you should experience and unexpected failure once in every million hours, assuming that you perform maintenence and replace the drive at the intervals recommended by the manufacturer. It does not mean that the drive will last for a million hours. It won't.

    By all means, stick with your hard drives - at some point, everybody will politely chuckle under their breath and point to the poor luddite who won't change.

  64. Re:Hell Yeah! by Aardpig · · Score: 1

    RTFA. 16 gigabit, not gigabyte.

    --
    Tubal-Cain smokes the white owl.
  65. Everything in it's Place. by twitter · · Score: 1
    How is a 2GB flash drive with only 100,000 cycles supposed to rival a much faster 500GB hard drive with a much, much longer life span?

    It's bigger than that, it's not going to rival something so big, it's going to be faster than most drives and you are so silly I wonder if you are that way intentionally. Let's quote the article:

    For mobile PCs - particularly thin-and-light models that do not require the larger hard drive capacities - the technology could extend battery life because solid-state Flash designs would be far more efficient than hard disks. Hardware design, durability and performance - particularly the boot-up sequence - would also be improved.

    Hard drives are not ideal in laptops. They are mechanically fragile and break when you drop them. The quickest way to extend battery life is to spin down your drive, but there are only a limited number of times you can spin them up before you burn the motor out. Laptop hard drives are typically slow, loud, hot and suck power. Most of my laptops have drives less than 16 GB, so I'd never see the difference.

    The techniques to extend the life of flash drives are well known. You put files that are rewritten often onto a RAM disk. That's how palm computers work. My Handspring gets by with 4MB of memory and is a useful tool. My Zaurus does much more with it's 64MB and a 512MB SD flash card.

    When your OS has adequate networking, you don't need to lug around all 500 GB of your music, photos and movies. If your OS has a 12 minute half life on any network, you might lug around an external drive. I leave mine at home tied to a cable modem. It's easier to sync that with everything.

    The best laptop is just big enough for a keyboard, screen and network hookups and weighs less than three pounds. It's just fast enough to run your stuff without burning your lap and it does not have sound like a vacuum cleaner. It runs for at least four hours without a charge, ideally twelve, so you don't have to lug around a power brick. Flash drives are part of that laptop.

    They are attractive as upgrades to my current laptops. At the predicted price of $90, it's cheaper than most laptop drives on the market. I'll be happy to drop half a pound of noise, heat and power drain from my current laptops when their eight year old drives start to fail.

    --

    Friends don't help friends install M$ junk.

  66. Re:Hell Yeah! by Aardpig · · Score: 1

    It was a fucking joke. Using flash as swap is as good idea as running NFS over dial up.

    --
    Tubal-Cain smokes the white owl.
  67. Re:OFFS! This is stupid. by Anonymous Coward · · Score: 0

    Digg.com called, they want their "conversation" back.

    That alone deserves insightful moderation.

  68. NO WAY IN HELL by Anonymous Coward · · Score: 1, Informative

    5 million writes? that might be true with the 180nm technology (actually it's not, with 180 it's more like 100k), but with samsung and toshiba pushing to 55nm this year and 45 soon after, they can't even guarantee more than 50k.

    worse yet, flash retention is going down too because of the smaller gates and use of multi-level cells. one electron jumping off the gate when you have seventy million of them in your huge gate is no problem, but when you are only holding a few dozen of them things in your tiny little gate, it's a much worse scenario.

    you either work for the company you advertise, or is just a fanboy. no way 16Gbit chips can have this much rewrite capability.

  69. Re:OFFS! This is stupid. by TheClam · · Score: 1

    First: "Just 5 years ago that amount of storage capacity was SciFi movie stuff of the future."

    And then: "I do not believe for a minute that NV memory drives will ever replace mag drives."

    Looks like someone prefers to be surprised. :)

  70. in some products [sometime]? by katepwa · · Score: 2, Interesting

    6000 networked desktops at my place of work, Windows XP notwithstanding, almost all 40G+ hard drives, most probably 90% empty, >200,000G of empty disk, bumps up the $/useful G somewhat, must be some market for 4-8G "drives" right now [again]!

    1. Re:in some products [sometime]? by HaydnH · · Score: 1

      I still use 18GB SCSI's in a raid 5 array for my home use, if I need more space I'll shove in another HDD... it's faster than having a single large HDD and add's a level of redundancy for what essentially is a low MTBF component. I currently have 4 disks, 2 of which I got off e-bay for £10 so the cost is dirt cheap also in comparison.

      --
      Time is an illusion. Lunchtime doubly so. - Douglas Adams
  71. Power consumption by phorm · · Score: 1

    That's another notch off the old power meter as well. I'm fairly sure flash-based media uses a lot less than the spinning of drive heads etc on a standard hard-disk.

    With all the advancements in computing that tend to require more power, that's a nice change. Especially since my mini-ITX system currently uses around max 35-40W already, including drives... this would likely be less if I used flash drives (an quieter).

    1. Re:Power consumption by Rakishi · · Score: 1

      HArd drives are far from the main source of power drain in a computer, and they're not the part which is causing the ever rising power requirements. Then again, those req. are only if you go for the uber comps, you can do quite well (and cheaply) with under 150W (without the crappy performence of mini-ITX).

  72. Re:Flash is a complementary technology, not a riva by jambarama · · Score: 1

    I'm not sure your analogy holds. These sorts of flash drives would be beautiful in portable devices, but I can see a terrific value using them in laptops. Imagine putting 5 x 16gb flash drives in a laptop in a RAID 5 array, giving 64GB usable space (more than enough for most laptops). You'd have a few huge advantages.

    1. You'd save a terrific amount of power (spinning CDs and HDs kills batteries faster than about anything).
    2. In a RAID 5 array (so long as only 1 drive fails at a time), if you have a drive failure, you could have a little program popup and say "look part of your HD has died. You haven't lost any data, so don't worry, but get this serviced immediately."
    3. With the RAID 5 controlled by the computer, the read/write times would be unreal. Fastest loading ever. 10 second boots and instant program loading!
    4. With solid state memory, no more hard drive heads scratching the hard drive platters if you drop your laptop.
    5. Flash memory is small. Even 5 chips would be much smaller than a 2.5 inch hard drive. More importantly, it is flat. So you could get laptops half inch thick (assuming a slot loading CD drive).

    I'm sure there are other advantages too, but this is what came off the top of my head. If we could get read/write speeds on flash drives up to the speeds DDR RAM has and you could have a computer that, when unplugged unexpectedly, doesn't lose anything it wasn't writing at that very millisecond. Boy wouldn't that be a leap ahead. (Windows Vista has an optional feature that uses flash drives as memory. Not as good as real RAM but leaps faster than virtual memory).

  73. Re:Flash is a complementary technology, not a riva by jambarama · · Score: 1

    To be honest I can only see Apple implementing something like that now, but eventually it'd hit the PC market. HP still innovates even if Dell doesn't.

    EDIT PARENT: I'm sure there are other advantages too, but this is what came off the top of my head. If we could get read/write speeds on flash drives up to the speeds DDR RAM has you could have a computer that, when unplugged unexpectedly, doesn't lose anything it wasn't writing at that very millisecond. Boy wouldn't that be a leap ahead. (Windows Vista has an optional feature that uses flash drives as memory. Not as good as real RAM but leaps faster than virtual memory).

    (I removed an "AND" that made the paragraph hard to read)

  74. Re:OFFS! This is stupid. by jizmonkey · · Score: 1
    Flash memory may have an indefinite SHELF lifespan

    Not with data on them they don't. Have you read a spec sheet? I think I saw a quote for 10 years last time I looked.

    --
    With great power comes great fan noise.
  75. Chips Vs media by phorm · · Score: 1

    Well, I know the storage media go higher than this. For work I have a Sandisk Cruzer 4GB. However, I'm guessing if I looked inside there might be a severak individual chips (say 4, 8-gigabit/1GB chips). If they could store 16gigabit/2GB on each chip, that's half the chips one would need and either 2x storage for the same size or less size needed.

  76. Flash is NOT slow! by Anonymous Coward · · Score: 0

    I'd bet that if you ever measured your actual real-world throughput from your hard drive, you'd be sorely disappointed. Just because it has a max outer zone transfer rate of 60-90MB/sec (I'll assume you aren't one of those idiots who thinks that because you are using SATA-II which is 300MB/sec that you are getting 300MB/sec for all transfers) doesn't mean it really gets that. In fact, you are often doing well to get 10% of that value, once you include the massive overhead of taking 5-10ms every time you seek.

    However, let's say whatever you do does mostly sequential transfers so that's all you care about. If you are using flash to replace hard drives, you are probably using more than one of those 16Gbit (2GB) chips. Even if each one is a painfully slow 5MB/sec (and newer ones are closer to 10-20MB/sec) you can run them in parallel. If you used 16 in parallel to create a 32GB drive, it is faster than even the fastest 15Krpm drive, probably faster than two of them striped together.

    Yes, its pricey, if those 16Gbit chips were $60/ea (just pulling a number out of my ass) 32GB would be nearly $1000. However, that's really not much more than two 15Krpm SCSI drives, and its far faster, lower power and more reliable. Whatever cost they sell 16Gbit chips for they could probably sell 8Gbit chips for half that, and 4Gbit for half that again. So maybe you don't need 32GB if you just wanted a fast boot drive. 8GB is more than enough for a boot drive for Windows (including Vista I'm guessing) and even the biggest Linux distros. If those 4Gbit chips were $15/ea, that'd be about the same speed for under $300. Sounds like a better deal that those 10Krpm Raptors people swear by, at least for your boot drive and your most important applications.

  77. Comparison by phorm · · Score: 1

    ATA hard disk:

    a) Low-medium power consumption
    b) Noise
    c) Heat
    d) Varying reliability, but at most I'd say medium as I never trust something important to just one disk
    e) High capacity
    f) Medium-High speed
    h) Medium-Large size (dimensions)
    i) Good price /unit storage, reliability (RAID, etc) or increased speed at to this, though

    Flash Drive:

    a) Minimal-lower power consumption
    b) No noise. Zilch
    c) Minimum heat if any
    d) Some brands have excellent reliability, depending on how you watch your write cycles
    e) Low comparative capacity
    f) Speed varies by operation (quite fast read, less so for write, dependng on random vs seq operation)
    h) Small size but this could be parallel to capacity
    i) Currently higher price /unit storage

    So the biggest obvious advantages of flash are really in the size, power use, and noise. Anything where you want one of the three to be low, you might go flash unless you're more concerned about capacity or # writes (say for swap).

    1. Re:Comparison by Arlet · · Score: 1

      It doesn't really make sense to use flash for swap space. Instead, just add more RAM which has a higher speed, and higher density per device.

    2. Re:Comparison by maxwell+demon · · Score: 1

      I think there's another advantage for flash: Better shock resistance.

      --
      The Tao of math: The numbers you can count are not the real numbers.
  78. Not necessarily true by phorm · · Score: 1

    As they become more widespread, dampening for such things could be included. Right now there's not much space for it. While it might take more juice to frag a hard-drive, I have seen IDE disks get zipped just as nicely. With the controller card gone, you have a disk full of wonderful data that you... can't use (unless you have a disk for the same chipboard to transplant).

    I'm not sure which handles better against magnets, etc, too (is flash succeptible?)

  79. but slowwwwww by sonictheboom · · Score: 1

    current flash cards are very very slow. ever try one out? booting takes 2 to 3 times as long.

  80. flash raid? by Anonymous Coward · · Score: 0

    i assume the boards that flash chips are mounted on already do this, but if you have say 4 flash chips making up a single 8 GB flash module, why not treat it like a raid and stripe data across the 4 chips? if you can only write 10 MB/s to each flash chip, striping the data across 4 chips instantly gets you 40 MB/s read/write performance for any block-level action - and most every os will read in block size increments, say 2 KB or 4 KB at a time.

    40 MB/s and no seek times? sounds good to me.

    sure that'd require more circuitry on the pcb the chips are on, but that's what mass production is for. if 4 striped chips is how things currently work, move to 16 striped chips of smaller dimensions.

  81. Re:Hell Yeah! by EvanED · · Score: 1

    Hey, swapping to NFS... now there's an idea...

  82. Re:Hell Yeah! by alc6379 · · Score: 1
    It was a fucking joke. Using flash as swap is as good idea as running NFS over dial up.

    My

    NFS

    root

    is

    mounted via

    dialup. What'

    s

    so wrong ab

    out that? OH NO! Office mate's got to make a phone call! Sure hope this unmounts clea+++NO CARRIER

    --
    I don't moderate anymore. Karma penalty for 90% fair mods? Can I mod that unfair?
  83. Great for... by pojo · · Score: 1

    Seems like this is perfect for car PCs. No moving parts, adequate storage for those nav systems and MP3 collections.

  84. Ah, but what if... by Anonymous Coward · · Score: 0

    your drive is almost full? In that case you're limiting writes/rewrites to a smaller portion of the disk. Your estimate assumes that writes/rewrites are spread uniformly when they are/cannot be spread anywhere near uniformly in the real world. In other words: Your estimate is wildly optimistic.

  85. Re:OFFS! This is stupid. by giorgosts · · Score: 1

    How about operating system and programmes on the flash and data on the hard drive?

  86. flash mem helps by groovy.ambuj · · Score: 1

    not of course by increasing storage but by incresing MTBF. yes no moving parts => much higher MTBF no moving parts => no seek required => dont incuss 6 mS for a 512B transfer from hard drive plan is to use small flash in conjungtion with regular hard drive and your flash contains everything need to boot your computer. believe me, you are going to love it when you can swtich it on (even windows!) in less than 1 sec.

    --
    This sig doesnt exist.
    1. Re:flash mem helps by Anonymous Coward · · Score: 0

      I am not going to believe you until you present the hard figures showing that your flash chip is faster than a modern HD (both for reading and writing) and has a higher MTBF than a HD (even for large numbers of updates).

      "your gut feeling" does not count, present the evidence please!

  87. Re:OFFS! This is stupid. by matt21811 · · Score: 1

    I have looked into the performance trends of flash vs hard disk and I suspect, in a just few years, that flash will catch up to the point that there is no meaningful difference in the sustained read and write speeds between the two. This means when the $ per meg is the same, there will be no logical reason to chose disk over flash, including in a hybrid setup. Could be a few years yet for desktops, 11 years for 3.5" IDE drives according to my study. Probably a lot sooner for 2.5" drives.

    I've linked to it elsewhere here but here is the link to my study again:
    http://www.mattscomputertrends.com/flashvsharddisk .html

  88. Re:OFFS! This is stupid. by Alioth · · Score: 1

    I wouldn't use that Maxtor as a video recording device on my Dad's racing motorcycle though - it wouldn't even last to the bottom of Bray Hill on the Isle of Man TT circuit. The only practical solution at the moment is tape (and it worries me what a battering that helical scan head on my portable video recorder is getting on road circuits). I'd much rather have a solid state video recorder than tape - especially if the recorder can be made smaller.

  89. Real gigabits or marketing gigabits? by Ezza · · Score: 1

    The article (as much as it confuses "Gb" with "GB") doesn't mention if it's actual computer gigabits ie. 16 x 2^30 or if it's the gigabits the marketing people like to use ie. 16 x 10^9.
    Marketing people like the later because it makes things sound bigger than they actually are.
    I stand by my theory that all marketroids harbour great feelings of inadequacy (probably for good reason, ie a small hard disk..).

    --
    I'm a perfectionist but I'm trying to cut back.
  90. I know the diff -- and I missed it in the article. by CFD339 · · Score: 2, Informative

    For me it was very helpful to see it posted -- I'm very much aware of the differences but missed the word gigabit in the article when I quickly read it. Reading that post was in fact "inciteful" to me and if I'd had mod points at the moment I'd have marked it so.

    --
    The problem with quotes on the internet, is that nobody bothers to check their veracity. -- Abraham Lincoln
  91. Re:I'm not surprised by JTL21 · · Score: 1

    Then don't bloody use it for paging, just use enough RAM that you don't need to page anything out.

  92. Oh damn. I did have them... by CFD339 · · Score: 1

    ....Didn't see that as I replied. Now they're useless in this thread. Damn my eyes.

    --
    The problem with quotes on the internet, is that nobody bothers to check their veracity. -- Abraham Lincoln
  93. Re:OFFS! This is stupid. by NickFortune · · Score: 1
    Wear leveling algorithms can make the write limit of flash irrelevant.

    Well, probably not if you have a file system that records access times. Of course you can turn that off.

    On the other hand, doesn't flash burn out much faster if it's being constantly written to? It'd make a bad coice for log and swap files, wear leveling or no.

    It's an exciting area, but I doubt I'll be an early adopter for the first wave of flash notebooks

    --
    Don't let THEM immanentize the Eschaton!
  94. Gordon by coofercat · · Score: 3, Funny

    "Flash! Flash! I love you, but we only have 14 hours to save the earth!"

    1. Re:Gordon by Anonymous Coward · · Score: 0

      dum dum dum dum flash aaa-a

    2. Re:Gordon by TheLetterPsy · · Score: 1

      He's a miracle!!

  95. hmm by Anonymous Coward · · Score: 0

    I think he was actually trying to say that flash cards aren't yet big enough to hold a complete windows folder! He tried to install everything to a 80GB drive but he still filled up his 9GB windows drive. Poor guy, he doesn't know what he is talking about....or is drunk - with this being slashdot, probably both.

  96. c00l.. 1337 :) by Anonymous Coward · · Score: 0

    now how mats pron dis holdz?

  97. HD replacement by VincenzoRomano · · Score: 1

    Speed, power hunger and heat seem to give non-volatile memories a plus over HDs.
    But for a real replacement of HDs in our PCs we'd need at least:
    1. Good "specific capacity", that is the ratio between memory room and physical volume, comparable with HDs.
    2. Medium durability. We use to both read and write on HDs. Current non-volatile memories seem to suffer a little bit when frequently updated.

    While point #1 seems to become more close every day, point #2 is the weakest one. But maybe someone will package some extra memory to automatically patch the broken banks (or sectors if you prefer).

    --
    Maybe Computers will never be as intelligent as Humans.
    For sure they won't ever become so stupid. [VR-1988]
  98. Re:OFFS! This is stupid. by matt21811 · · Score: 1

    "doesn't flash burn out much faster if it's being constantly written to? "
    This is the first time I have ever heard this idea. If you have a source, I am keen to learn from it. It will modify my opinion.

    I think that people have a poor perspective on just how long a flash drive can last with wear leveling.
    Taking a the example of a theoretical hard disk replacement flash drive of 200GB with a write speed of 40 Megabytes per second and doing some basic calculations shows that it could be written to continuously for just over 15 years before every block passed the 100 000 write mark. The equivalent of todays 200GB drive some 15 years ago was the 210MB disk. There are not many machines running today with 210MB hard drives, let alone dong the kind of work that requires continuous writing to the disk. And 100 000 writes is often considered a minimum. Average failure figures are often quoted as 1 million writes. Worrying about it wearing out just isnt worth the time.

    " I doubt I'll be an early adopter for the first wave of flash notebooks"
    Wise advice for any technological advancement.

  99. Flash will not replace HDs... yet. by NimbleSquirrel · · Score: 1
    While it is nice to think that we will be seeing a NAND Flash replacement to hard drives (especially in laptops), I don't think we will be seeing them any time soon. Even then they will still remain very expensive. As has been pointed out earlier, these are 16Gb chips, not 16GB, so you will need 40 of them to match the storage capacity of an 80GB HD. Flash behaves very differently in write cycles than HDs. Once a byte is programmed, it can't be rewitten untile the entire block has been erased, so you will need a decent ammount of cache to make up for that. On top of that, Flash has a limited number of erase/write cycles. Virtual memory and/or swap partitions (not to mention constant writing and rewriting of data during normal use) will very quickly kill your expensive NAND Flash drive. Flash is a good stroage medium, but is simply no effective replacement for HDs yet.

    In saying that, there are companies that do make Flash based solid state disks that solve most of these problems. I'd say that even the finite write/erase life is made less of an issue by large ammounts of caching and something akin to RAID 6. Since those manufacturers supply for military applications, be prepared to pay massive ammounts for it.

    I can't see the cost of producing NAND Flash chips (even at 16Gb) in high enough quantities to equal the cost per GB of HDs. Once the cost drops further, packaging enough of them for redundancy into a 2.5" form factor with all the caching other circuitry (most likely some kind of hardware RAID 5 or 6 controller) will start to make economic sense. Until then I think we are stuck writing magnetic bits onto spinning chunks of metal. ;)

    1. Re:Flash will not replace HDs... yet. by Bassman59 · · Score: 1

      "On top of that, Flash has a limited number of erase/write cycles. Virtual memory and/or swap partitions (not to mention constant writing and rewriting of data during normal use) will very quickly kill your expensive NAND Flash drive. ... In saying that, there are companies that do make Flash based solid state disks that solve most of these problems. I'd say that even the finite write/erase life is made less of an issue by large ammounts of caching and something akin to RAID 6."

      At my last job, I designed a NAND flash drive that mounts in a PMC site, so I know a little bit about this. Nope, no large amounts of caching or RAID was used to mitigate the wearout problem, although we considered putting some SRAM on it to use as a cache to speed up access.

      Flash file systems use wear-leveling algorithms to help mitigate the finite lifetime of the flash. Basically, you never keep writing to the same location over and over again. Rather, you always write to new locations. This is especially useful when you read, update and write the same file over and over again, for example if you were using the "disk" for a swap file. The good news is that the low-level file system takes care of these details, so application programmers don't need to worry about it. With a large "disk" the file system can really spread the love and increase the disk lifetime.

      Using a flash disk for a swap file might be painfully slow, but "painfully slow" is application dependent.

      The other thing to note is that your flash file system needs to take care of bad blocks. When the chips are tested at the factory, any bad blocks are noted in the memory "spare area." The file system needs to read this bad-block info when the "disk" is initialized and it has to keep track of (and not use) the bad blocks. Now, when doing writes or erases, the file system needs to verify that the operation was successful, and if a bit in a page doesn't verify, it has to mark the entire block as bad. You lose 16 k bits if one bit fails.

      Flash is the thing to use when your applicaton can't tolerate the power consumption of a spinning drive, or if the environment precludes the use of a mechanical drive. You pay for this in terms of much smaller storage capacity.

      -a

  100. Re:OFFS! This is stupid. by NickFortune · · Score: 1
    This is the first time I have ever heard this idea. If you have a source, I am keen to learn from it. It will modify my opinion.

    That was according to a colleague when we were investigating flash as a candidate for a rolling network recovery buffer. I just asked him about the source, and he says it came from an intel or possible microsoft research paper.

    Sadly he couldn't find the link, and nor could I after a half hour of googling, which is all I can really allocate out of work time.

    I might have another look tonight. I must admit, I'd like to read this paper myself.

    --
    Don't let THEM immanentize the Eschaton!
  101. Re:Hell Yeah! by Anonymous Coward · · Score: 0

    From www.microsoft.com/vista

    External memory devices

    Adding system memory (RAM) is often the best way to improve your PC's performance. More memory means more applications are ready to run without accessing the hard drive. However, upgrading memory is not always easy. You must learn what type of memory you need, purchase the memory, and open your computer to install the memory--which sometimes can invalidate your support agreement. Also, some machines have limited memory expansion capabilities, preventing you from adding RAM even if you are willing to do so.
    Windows Vista introduces a new concept in adding memory to a system. USB flash drives can be used as External Memory Devices (EMDs) to extend system memory and improve performance without opening the box. Your computer is able to access memory from an EMD device much more quickly than it can access data on the hard drive, boosting system performance. When combined with SuperFetch technology, this can help drive impressive improvement in system responsiveness.
    EMD technology is both reliable and secure. You can remove an EMD at any time without any loss of data or negative impact to the system; however, if you remove the EMD, your performance returns to the level you experienced without the device. Wear on the USB drive is not an issue when using it as an EMD. A unique algorithm optimizes wear patterns, so that a USB device can run as an EMD for many years, even when heavily used. Finally, data on the EMD is encrypted to help prevent inappropriate access to data when the device is removed.

  102. Mac related rumours? by aallan · · Score: 1

    The rumours in the Mac community are hinting that key products were missing from the Steve's keynote speech due to supply issues with the new Intel Core Duo chips. I'm starting to winder whether it wasn't the Core Duo chips that were in short supply. There were predictions and rumours of "instant-on" hard drive less Macs based on flash chips in the run up to Macworld. Perhaps we still have "One more thing..." still to come?

    It'd certainly fit with my own speculations about the Mac Pro line...

    Al.
    --
    The Daily ACK - Eclectic posts by yet another hacker
  103. Re:Hell Yeah! by MadCow42 · · Score: 1

    Yes, and you can re-install your OS and all your apps (and data) each time you shut down, reboot, or have a power failure. (RAM doesn't hold data without power, silly!)

    Better not run Windows... you'll be rebooting/reinstalling all the time. At least with Linux you stand a chance of a running system for a few months (or years...) at a time.

    MadCow.

    --
    I used to have a sig, but I set it free and it never came back.
  104. Re:Flash is a complementary technology, not a riva by codemangler · · Score: 1

    Hard disks may be physically larger and slower for random access, but they are faster than Flash for large sequential reads, much in the same way that the hare is faster than the turtle in that old fable.

    As I'm sure you know, the turtle was faster:

    "A HARE one day ridiculed the short feet and slow pace of the Tortoise, who replied, laughing: "Though you be swift as the wind, I will beat you in a race." The Hare, believing her assertion to be simply impossible, assented to the proposal; and they agreed that the Fox should choose the course and fix the goal. On the day appointed for the race the two started together. The Tortoise never for a moment stopped, but went on with a slow but steady pace straight to the end of the course. The Hare, lying down by the wayside, fell fast asleep. At last waking up, and moving as fast as he could, he saw the Tortoise had reached the goal, and was comfortably dozing after her fatigue.

    "Slow but steady wins the race."

  105. One word by Huh? · · Score: 1

    Power.

  106. whats XPe like? by petermgreen · · Score: 1

    does it miss features that home/pro have?
    does it cost a lot to license? (compared to what white box vendors pay for oem versions of home/pro)
    can licenses only be bought in bulk?

    --
    note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
    1. Re:whats XPe like? by bynary · · Score: 1

      You're missing the point. XPe offers OEM manufacturers the ability to install whatever components of XP they want. From experience, the installer allows you to hand pick which components get installed and which ones don't, unlike the Pro/Home installs that install everything. It reminds me more of going through a Linux install. You only have to install the minimum system components necessary to get the job done. That's why it's marketed as XP Embedded. You can whittle the size down pretty small. I think the install I saw being done of XPe fit in 128 MB (that included a custom GUI and use of part of the 128MB as system RAM). I know that's not very impressive for you Linux people out there, but that's pretty impressive from a Windows XP standpoint.

      --
      http://www.bynarystudio.com
    2. Re:whats XPe like? by mythosaz · · Score: 1

      Also, to be clear, it's not just like leaving off Calculator and Wordpad. You can, for example, leave off support for FAT if you're NTFS only - leave out all networking components - leave out the GUI if you want.

      When you build with XPe, you simply pick an item you want, and dependancies are resolved. If you select HTML Help, for example, you'll end up with an entire build that does everything necessary to run HTML Help (including a GUI and file system) but does nearly nothing else.

      You can then go and trim some of the fat from the "necessary" components.

    3. Re:whats XPe like? by petermgreen · · Score: 1

      all sounds nice but that still leaves two questions.

      is XPe expensive and/or only purchaseable in large bulk (and therefore only an option for large manufaturers).

      if you pick everything on XPe do you end up with everything that home has? everything that pro has? less than either? more than either?

      --
      note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
    4. Re:whats XPe like? by mythosaz · · Score: 1

      You may never see this reply. Perhaps you check your home-page.

      If you pick every conceivable option in XPe, you end up with XP Professional with all the bells and whistles. You'd get Home or MCE or Tablet, but those variances are determined by your product key. (EG. Tablet is Home/Pro with an additional CAB installed, and installing that CAB or not is determined by your product key. Ditto for Media Center Edition).

      To the best of my knowledge, XPe isn't sold a la carte, and you can't buy a license from NewEgg. You might have to be a ODM or OEM. You can 120-day-trial XPe as a download from Microsoft. You need their download and possibly your XP disc -- although, honestly, I think you get it ALL from the download (IIRC). You can then build your very own XPe machine. You'll just need to supply drivers for the hardware that you're putting it on - as there's no setup process on first-run.

      Enjoy.

  107. Bull by Anonymous Coward · · Score: 0

    RAM is measured in bytes not bits. It's MB and GB not Mb and Gb.

    1. Re:Bull by be-fan · · Score: 1

      RAM modules are measured in bits you nimrod.

      --
      A deep unwavering belief is a sure sign you're missing something...
  108. Card not booting cam - Data scrambled not lost.. by bdwoolman · · Score: 1

    It was with an older camera. If I was deleting snaps and the power went off (auto power save shutdown) then the card would not reboot in the camera in many cases. To read the remaining pictures I had to use an adapter -- PCMCIA or USB both worked. Remember this: if your card malfunctions in the camera then all is NOT lost. The FAT table on the card was screwed and pictures were all out of order. Sometimes they had duplicate names! It was messy, but not a disaster as no pictures were lost. Once I figured out the pattern I disabled the power save function if I was deleting lots of pictures. Problem solved.

    After snagging the data I could reformat under windows using vanilla floppy FAT, which is the format these cards usually use. No need to worry much. Newer cameras I have had did not have this problem since the cards were big enough not to delete on the fly. Still, if I have to do a rare major delete marathon on my new camera I disable auto off, but only out of paranoia. But the damage I report here is the kind of thing that happens if you pull your card without unmounting it. Anyway no big deal to have your CF card go wonky, but a pain -- especially if it is on Mars. They reformatted and it booted fine. Rovers have kicked ass ever since.

    d:-b

    --
    "No fear. No envy. No meanness." Liam Clancy
  109. Thanks. That was interesting by bdwoolman · · Score: 1

    Always like to learn a neat detail. I knew they rebooted and solved the problem. Cheers, b

    --
    "No fear. No envy. No meanness." Liam Clancy
  110. Agree my problems with flash were software based. by bdwoolman · · Score: 1
    And on mars, too. I still like flash for a mobile player that stands up to a workout. The idea of bouncing around on a run or a ski trail with a HDD in my pocket gives me the willies. People seem to be happy though, so I know it is my own HDD paranoia. But there is a place for the Shuffle and the like. And it is true that tiny HDDs are cool as bees as on rollerskates. Fact is all this stuff is great. 500 pictures on a card, your music collection in a package the size of a cigaret box. Whoa. Now, Where is my robot? My office is a mess. Cheers,

    d:-b

    --
    "No fear. No envy. No meanness." Liam Clancy
  111. Re:Flash is a complementary technology, not a riva by Anonymous Coward · · Score: 0

    I'm not sure your analogy holds.

    ...
    ...
    ... Ya' think?

    Go back and read his username. Also, jet engines don't have propellers -- hence the joke.

  112. everything get firkled if you firkle with it! by Anonymous Coward · · Score: 0

    Doesn't the fact that the Mars Rover used flash suggest that flash is a really solid technology?

    I think your hard drive will be upset if you mess with the power during write too. Not to mention your PC.

    It still amazes me that hard drives are as reliable as they are, with all those moving parts. Flash doesn't have much to break, by comparison.

  113. called worst idea of 2006 by peter303 · · Score: 1

    The current issue of IEEE Spectrum (the electrical engineers society news monthly) called flash hard drives one of the top ten worst ideas of 2006. First, they say the is over a hundred times higher- $45 / GB for flash versus $0.50 / GB for a commodity disk drive. (People estimate Apple is only paying $10-$20 / GB for nano-pods.) Second, bulk transfer speeds are several times slower for flash than disk. Third, flash can only be written about 100,000 times before it burns out, while disk can be written billions of times.

    (DISCLAIMER- not necessarily my opinion.)

    1. Re:called worst idea of 2006 by Anonymous Coward · · Score: 0

      Yeah, just like LCD monitors were a horribly stupid idea when they first came out, because they cost several times more than CRTs and didn't have nearly the same picture quality. Or how optical drives were vastly inferior to floppy drives because they were read only. Why waste time developing such obviously deficient technologies?

      It's funny, I would have expected a magazine published by IEEE to be less frustratingly idiotic.

    2. Re:called worst idea of 2006 by Anonymous Coward · · Score: 0

      And what is the last portable computer you remember with a CRT-Display ?

      The Osborne comes to mind ...

      If you spend $3000 on a one of the fancier "1Kg" notepads, you probably would not blink at an extra $200 for 10 .. 20Gig of a solid state disk, if that gives you

      - 10% longer runtime
      - 10% less weight
      - Several times higher chance of retrieving your data if you drop the notebook

      It is important to keep in mind that all current OSs have been designed to make the most of the severely limited hard-disks.

      Once both OS-drivers and Nand devices have been optimised for solid state storage, nobody except video users will want to look back to harddisks.

      Here is something to think about:

      Current Nand-devices read a 2K sector in about 10us from the Memory array into the read latches.
      This is a raw read performance of 200Mb/s.

      Nothing (except, until now, lack of demand) prevents Samsung from putting a SATA Interface on the nand flash chip.
      Then, a *single* nand flash beats your top-of-the-line Harddisk for read speed (=booting).

      Now put 8 of these these in parallel ..

      Nandflash has the promise to improve on two of thoughest issues with nonvolatile storage in the last 50 years:

      - reduce seek-time by three orders of magnitude (10ms -> 10us)
      - increase read-bandwith by one order of magnitude, up to maybe 1/10th to 1/2 of ram bandwith

      Think for example what happens if there is a page-fault for code and the page is not cached :

      hard-disk:
      do a task-switch, then switch back after 10ms.
      Even if there is a another process that can use the cpu, you lose 1Mb of cache twice.
      Cost of reloading half of it : 2x0.5Mb = 1Mb at 5Gb/sec = 200us

      Nand-Flash:
      You can read a 4k page in 50us, so it is cheaper to wait than doing a taskswitch !

  114. Re:OFFS! This is stupid. by Savantissimo · · Score: 1

    According to Tom's Hardware, as of August 2005, the fastest flash on the market, Memina's 4GB, has 30MB/s read and 24MB/s write - assuming it's compatible with your system, which it often isn't. Yes, if someone created a RAID controller for the things, this could go to 100MB/s or more, but that isn't available yet. The access times are great, though, on the better brands - 0.5 ms for the Kingston Elite 2GB or 0.7 ms for the Memina. Crap brands can have ridiculous access times, though - nearly 30 ms.

    --
    "Is life so dear, or peace so sweet, as to be purchased at the price of chains and slavery?" - Patrick Henry
  115. Re:OFFS! This is stupid. by kabocox · · Score: 1

    The real story here is that flash is trouncing disk in improvements in Megs per $ and will one day catch up to and overtake disk. And it will be sooner than mmost people expect.

    I don't see that. I think that we will see 2-4 TB HD when in the next 10 years. I'd be happy if flash caught up, but I'd think that flash would say get to be affordable on in the 500 GB range unless some prices were lowered to make the product very attractive. I think that both techs have a place. I could easily see a system with lets say 16-32 GB of RAM with a 100 GB flash drive and several 1 TB setup with RAID storing next gen. HD TV or DVD. The humans shouldn't really decide which storage medium needs to be used at any given moment, Windox X or Linux should automatically take care of that for the users so that it just works.

  116. NAND-SSD is a loser by necro81 · · Score: 1

    In the latest issue of the IEEE Spectrum, the feature article is a list of some of the biggest winners and losers in recent technology. Samsung's play for solid state drives (SSD) breaking into the mainstream hard drive market was listed as one of the losers for several important reasons:

    * Significantly higher cost than magnetic drives (60-70 times the price per GB).
    * Signficantly smaller capacity tha magnetic drives (even a 16 gigabyte drive doesn't come close to the 80 gigabyte drives that are now becoming standard in a lot of laptops).
    * Slower read access (about 1.5x slower), and slower write access (2-4x slower).
    * Finite write cyles before the cells start crapping out (10^5 to 10^7).

    These are compelling reasons that don't seem to be going away any time soon.

  117. NAND-SSD a loser, II by necro81 · · Score: 1

    Sorry, forgot to include links:

    IEEE Spectrum magazine

    And the article describing NAND drives as a loser.

  118. Not necessarily. by emil · · Score: 1

    If you fill 95% of the device with some static file, then repeatedly write/erase some small file in that remaining 5%, you will probably cause that 5% area to fail much faster.

    Or is the wear leaving algorithm able to move static storage around the device also?

    1. Re:Not necessarily. by morcheeba · · Score: 1

      Bingo. I hadn't thought of that, but there is no reason the wear-levelling algorithm can't touch less-used blocks, so it's not a problem.

  119. Re:OFFS! This is stupid. by Martin+Blank · · Score: 1

    The last manufacturer that I think had a fairly universal bad reputation was Conner, and they were bought out years ago. A friend and I have had a serious problem over the years with WD; other friends swear by them. I prefer Maxtor for the most part, while others won't touch them. Seagate is about the only company where I've never had a drive fail, but others have had them bad out of the box.

    Most people here don't have enough experience with hard drives to have their failure biases mean anything statistically. I recognize that WD is a good brand, just that I've had bad luck with them.

    --
    You can never go home again... but I guess you can shop there.
  120. Everything get firkled if you firkle with it! by bdwoolman · · Score: 1

    True enough. Nothing against flash. Cannot wait for the near instant boot a flash based computer would have. b

    --
    "No fear. No envy. No meanness." Liam Clancy
  121. Re:Flash is a complementary technology, not a riva by zippthorne · · Score: 1

    slow but steady only wins races when fast isn't also steady.

    --
    Can you be Even More Awesome?!
  122. Limited rewrites for MIT's $100 laptop? by Peter+Trepan · · Score: 1

    The comments on this thread are the first I've heard of a flash drive's limited rewrite capacity. Does this mean MIT is about to supply the third-world with a bunch of laptops whose drives will need to be replaced every few years?

    --

    Step into a huge movement. Don't Tread In Me.

  123. That makes a bit more sense... by ashpool7 · · Score: 1

    ... if the driver reads the wear count and does the translation on the fly. This sounds like something that should be standardized, though. Because using it on a foreign computer where a generic SBP2 driver is loaded (and not the special one) would render it useless again.

    I could foresee a hybrid hard disk/flash disk, where the high read/low write areas are remapped to flash (say, OS files) and everything else is on disk. Hard drives already have something of this capability with their caching mechanism.

  124. be smart use SMART by nietsch · · Score: 1

    if you use smartd, it can warn you of disk that are about to break down. Disks themselves can correct bad sectors with a checksum per sector, if the sector is really gone it is moved offline and swapped for a spare sector. Similar methods might be used with flash.

    --
    This space is intentionally staring blankly at you
  125. I call. by maxume · · Score: 1

    You must have something else going on, C:\Windows + C:\Program Files\Common Files + C:\I386 (windows install stuff) is right around 4.5 gigabytes. That's with photoshop and office light installed. Not to mention a bunch of other shit.

    --
    Nerd rage is the funniest rage.
  126. Re:Flash is a complementary technology, not a riva by maxume · · Score: 1
    (put the OS, most used applications, and swap-space on it, and use traditional harddisks for your videos/music/porn).

    Does it really make sense to put swap on flash in leiu of say, adding more RAM? I actually am curious, I want some informed opinions, or at least more informed than mine.

    --
    Nerd rage is the funniest rage.
  127. Yuk by NiteShaed · · Score: 1

    >>vanilla floppy FAT

    That may be the most disgusting 3 word combination ever seen on Slashdot.

    --
    Some bring out the best in others, some the worst. Some bring out far more.
    1. Re:Yuk by bdwoolman · · Score: 1

      Thank you. I tries. I tries. b

      --
      "No fear. No envy. No meanness." Liam Clancy
  128. Re:In 40 years it will be a moot point by vertinox · · Score: 1

    I'm not saying I'm the expert on this -- I'd appreciate it if someone could explain to me that the flash memory will actually last 40 years or so. But I doubt it.

    In 40 years, you might not have an OS that is capable of reading the flash memory of today if you just pulled the flash memory out of a dusty bin.

    Heck... We might even done away with binary all together. We don't know and can't predict the future to an exact anymore because technology is changing at an accelerating pace.

    People with punch cards and old mainframe tape drives have the same problem today with getting and those were only 20 to 30 years ago. Even though there might not be any data loss on the medium they just don't have anything that can read it and transfer to a modern device.

    The best bet is that in 5 years from now you just transfer everything to a new medium and go with baby steps. Yes it is a pain, but in 5 years you'll probaly be able to store everything in an increasingly small but exponentially larger capacity.

    Like doing backups to floppy and then burning those floppies to CDR and then the cd images later to DVDRs and those to Blu-ray/HD-DVD and so on as the technology becomes available.

    --
    "I am the king of the Romans, and am superior to rules of grammar!"
    -Sigismund, Holy Roman Emperor (1368-1437)
  129. Solid-state disk cache? by stuie · · Score: 1

    I'm suprised that I've not seen anyone suggest this, but I can imagine a solid-state "cache" for a hard disk being very useful. For small writes (e.g. updating file access times), you needn't spin the disk up at all. When the cache fills above a certain threshold, spin the drive up, and write the data back. While doing really heavy work, the drive would bypass the solid-state cache entirely, so you'd get higher throughput. When doing light work, the disk can stay powered off for hours... and you'd have still high capacity.

    Okay, so it doesn't fix the problem that hard disks are quite fragile... but it should still reduce power consumption, and should cost less (at least for the short term).

    --
    Stuart Brady
  130. Rival Hard Drives? Not yet by InsaneProcessor · · Score: 0

    This is not yet a rival of hard drives. Cost per megabyte is still too high. Since we can buy 250Gbyte hard drives for around $100 the same in flash is about $11,000. It has a long way to be useful and cost effective as a HD.

    --

    Athiesm is a religion like not collecting stamps is a hobby.
  131. will we ever get fully integrated? by utexaspunk · · Score: 1

    So will we ever get computers where the processor, RAM, and flash memory storage are all on the same chip? it seems like you could make a very space-efficient computer that way, and probably produce the whole thing for less since you would have less individual parts and connectors, etc. Or would that cause heat dissipation problems? When do I get my tablet that is one solid block consisting of one chip, one battery, the display, and a few other minor components? (maybe with channels full of alcohol hollowed into the block?)

    1. Re:will we ever get fully integrated? by Bassman59 · · Score: 1
      "So will we ever get computers where the processor, RAM, and flash memory storage are all on the same chip?"

      No, for a couple of reasons. One is that flash and CPUs are built on very different IC processes and integrating the two is difficult. (This is why FPGAs with built-in configuration flash are always one or two generations behind their counterparts that use an external configuration EEPROM.)

      A second reason is that once you integrate all this stuff, you lose the ability to add more memory. Don't like 512 MB? Tough, that's all that the chip supports.

  132. Re:Intel does have something in the works. by vertinox · · Score: 1

    Intel has already demonstrated at NAND flash drive used to reduce boot times:

    http://www.pcworld.com/news/article/0,aid,123053,0 0.asp

    So you are probaly right that Apple and Intel might be cooperating on a similar project.

    I remember that Steve Jobs gave a key note a few years back on talking about how they wanted to reduce boot times to 0 because it would save people hours a year waiting for the computer to boot. However, I wouldn't expect anything til late this year or 2007.

    --
    "I am the king of the Romans, and am superior to rules of grammar!"
    -Sigismund, Holy Roman Emperor (1368-1437)
  133. pay attention by Anonymous Coward · · Score: 0

    We're talking about replacing hard drives with flash memory (not replacing DRAM with flash memory), and the result would be far faster.

  134. From TFA... by ThankfulJosh · · Score: 1

    "The cadmium atom that has lost an electron becomes a negatively charged ion, which can then be controlled with an electrical field," said Daniel Stick, a doctoral student in the University of Michigan's physics department who participated in the work.

    Any 7th graders out there want to point out the wrong physics in that statement?

  135. Ok now, think hard by p3d0 · · Score: 1

    Is there really no way that the article could be correct on this?

    --
    Patrick Doyle
    I mod down every jackass who puts his moderation policy in his sig. Oh, wait a sec....
  136. Re:Hell Yeah! by ePhil_One · · Score: 1
    RTFA. 16 gigabit, not gigabyte.

    Not sure how to respond because I don't know what you mean. I know the difference between gigabit and gigabyte, but I'm not sure how it relates to my concerns.

    Are you concerned that I mentioned the 4GB address limit and want to point out that 16Gb is less than 4 GB? I realize that, and also realise that multiple chips are often used together, and that his computer already had RAM in needed for running his web browser, etc. TFA was about these chips replacing hard drives, so staying on-topic those are two reasons why RAM is not a viable alternative.

    The OP stated he would use them for swap. This could have been a joke because flash does not do well in an extensive rewriting environment like swap, or he could have been serious because swap is one area that would see a lot of benefit from fast random I/O.

    The grandparent responded to this that he'd rather just use RAM. Now, you could thing he was arguing for using a "ramdisk" for swap (which is faster still, and doesn't suffer the re-write issues of flash), and my "data is lost when power is lost" is something of a non-issue in that use. Or he could have ben cracking a joke about using a RAMdisk for swap, the place you put stuff when you run out of RAM (ie its just a wacky idea, though some bad memory models made this a good idea in the past).

    Of course, TFA is also kind of silly because flash chips have long been used as a substitute for hard drives in systems. The Pix 515 is a BX based 166Mhz PC with its OS on a flash chip, just as an example, not to mention just about every PDA I've owned over the last 5 years. But its a press release you you sort of expect that.

    --
    You are in a maze of twisted little posts, all alike.
  137. Re:OFFS! This is stupid. by earnest+murderer · · Score: 1

    You're making the same mistake most people do. Flash drives are only going to be as fast as the MFR's implementation of the USB interface. Just as external hard disks are far faster than that bus (firewire or USB) permits. Toshiba's current generation, for instance, MLC NAND chips are rated at 108/45 MB per second. But you'll need a proper interface, which a USB device chain is not.

    --
    Platform advocacy is like choosing a favorite severely developmentally disabled child.
  138. number of writes by fufinache · · Score: 1
    >Of course there is also that 'number of writes' issue.

    Hard drives also are depend on moving parts which make them less accident resistant, give them a 'number of start/stop cycles' issue. They use magentic storage (data will fade out after long periods of time). I'd trust NAND memory in a laptop more than a hard drive anyday. I don't trust technology. I would buy a new hard drive every 5-6 years to replace an old one that could be working perfectly fine. (Usually means I get a lot more storage too, but that's coincidence)

  139. So stripe it by David+Gould · · Score: 1


      Flash-based drives aren't even up to UDMA66 speed yet. For notebooks, my 60G Hitachi 7200rpm drive will be faster than flash in every situation.

    At 16Gb (i.e., 2GB) per chip, a 60GB Flash "drive" would have to contain 30 of these chips, which could presumably be arranged in something like a RAID-0 striped array. So the question becomes... are those Flash I/O rates within a factor of thirty of UDMA66 speed yet?

    --
    David Gould
    main(i){putchar(340056100>>(i-1)*5&31|!!(i<6)<< 6)&&main(++i);}