Slashdot Mirror


New Serial ATA Standards Target SSDs, Tablets

crookedvulture writes "SATA-IO has devised a couple of new storage interfaces optimized for solid-state drives. To serve high-performance SSDs that are fast approaching the 6Gbps ceiling imposed by the current Serial ATA specification, the SATA Express standard will meld the Serial ATA software stack with PCI Express to offer up to 16Gbps of bandwidth. SATA Express isn't expected to be completed until the end of the year, but the new uSSD standard looks to be ready for prime time. Designed for tablets and ultraportables, uSSD sticks with current 6Gbps speeds but ditches traditional Serial ATA connectors, allowing SSD controller chips to be soldered directly to motherboards. SanDisk already has a 128GB uSSD primed for ultrabooks."

113 comments

  1. Come again? by LordLimecat · · Score: 1

    uSSD sticks with current 6Gbps speeds but ditches traditional Serial ATA connectors, allowing SSD controller chips to be soldered directly to motherboards.

    You best be joking.

    1. Re:Come again? by gilesjuk · · Score: 2

      It's for tablets. You simply don't have room inside for big bulky connectors that nobody is ever going to get access to.

    2. Re:Come again? by donkeyoverlord · · Score: 1

      Does this really need a standard? Seems like something that a manufacture could just do.

    3. Re:Come again? by Sir_Sri · · Score: 1

      Or if you're going to use an SSD as a cache on a regular mobo. I'm not suggesting that's a great idea, but I've seen one MOBO like that already (a gigabit z68).

    4. Re:Come again? by 0123456 · · Score: 1

      So your motherboard is toast once the SSD dies? Sounds like a great plan for motherboard manufacturers.

    5. Re:Come again? by copb.phoenix · · Score: 1

      If there's any takeaway from any market at all for milking cor the money, it's that if you can plan for obsolescence, you do it from the get go.

      I don't think there's any need to solder these things straight to the boards - I think it's a mutual arrangement between these major manufacturing companies to begin forcing hardware updates to happen faster and repairs to be impossible to have serviced by a private worker. Worse still, that to upgrade a single major part, you'll have to buy an entirely new machine.

      See: Apple. Notorious for leaving out things that make no sense or making upgrades that feel superficial (like adding the camera that should have been there to start with in the iPad 2, anyone?). Also, not interested in the flamefest usually associated with that statement.

    6. Re:Come again? by adisakp · · Score: 2

      uSSD sticks with current 6Gbps speeds but ditches traditional Serial ATA connectors, allowing SSD controller chips to be soldered directly to motherboards.

      You best be joking.

      MacBook Airs are flying off the shelve with RAM already soldered onto the MB. Soldering on the SSD allows a little more space (perhaps for more battery) or for even more weight savings.

    7. Re:Come again? by DemonGenius · · Score: 1

      Because there's not enough profit in allowing consumers to expand their hardware and delay having to buy yet another appliance at an exorbitant price.

    8. Re:Come again? by jedidiah · · Score: 1

      Big bulky connectors? When's the last time you looked at the back of a hard drive? 1990?

      Standard connectors aren't exactly 'big and bulky' now.

      I also question the necessity of a faster interconnect. Are drives really sustaining those kinds of speeds? A lot of the reviews seem to indicate that these drives aren't really all that. Regardless, even the full potential of current SATA interconnects are a vast improvement to spinny disks. Upgrades in storage capacity and improvements in cost per TB would be much more useful developments.

      Let me fill up my current array hardware with SSD first without requiring a second mortgage on the house. Then tweak the underlying hardware.

      Fondleslabs are nice and all but they still have meagre storage and probably should not be the drivers if new standards.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    9. Re:Come again? by tlhIngan · · Score: 3, Interesting

      Does this really need a standard? Seems like something that a manufacture could just do.

      They already have. SATA-over-mPCIe has been around since the original eeePC - the SATA SSD it uses was mounted in a mini-PCIe looking slot. But it wasn't, since it ran SATA signals over it.

      A more recent example started since the 2010 Macbook Airs which had a bog-stadanrd SATA based SSD in something that looked like a mini-PCIe slot - again, it was SATA signals wired to the slot.

      This spec just makes it official so everyone can build adapters, SSDs and laptops based on it and be standardized across the entire line. otherwise you'd have formfactor issues, possible pinout issues, etc.

    10. Re:Come again? by jedidiah · · Score: 1

      It does seem to fit into the domain of systems a that are completely non-user-servicable.

      So why does standardization even matter at that point?

      --
      A Pirate and a Puritan look the same on a balance sheet.
    11. Re:Come again? by symbolset · · Score: 1

      Onboard storage in tablets too.

      --
      Help stamp out iliturcy.
    12. Re:Come again? by Anonymous Coward · · Score: 1

      Have you tried putting a SATA cable inside a cell phone? It doesn't fit and a cell phone isn't far behind the tablet tech wise, open up a tablet or cell phone and you will see most of the cables are so small you need tweezers to disconnect them and the wire itself is plastic with metal silk screened on it. My only question is what is it about the current SATA standard that prevents a board manufacturer from skipping the connector and using traces. do the chips really care about the shape of the plastic near the wires? Or is this just some power/wire length/timing limits that are getting adjusted?

    13. Re:Come again? by jedidiah · · Score: 1

      The SATA cable isn't the problem.

      The SATA drive is the problem.

      If only there were some other form factor out there that was small enough for a phone...

      --
      A Pirate and a Puritan look the same on a balance sheet.
    14. Re:Come again? by fuzzyfuzzyfungus · · Score: 3, Interesting

      It would probably be most accurate to think of this "uSSD" as a faster, more PC-architecture-oriented version of the "eMMC" JEDEC standard for soldering flash directly onto a motherboard, with a lower board space, pin count, and controller requirement than raw flash chips.

      "eMMC", which is basically an MMC card's guts in a BGA package, is already quite popular in things like cellphones(ever wonder why some cellphones filesystem names suggest that they have an MMC card that they don't really? It's because they do, in software terms...) "uSSD" will, presumably, be the big brother of that standard, putting SATA signals and power over a standardized BGA arrangement, rather than using MMC signals and power...

    15. Re:Come again? by fuzzyfuzzyfungus · · Score: 2

      Manufacturers want to be able to switch flash suppliers without doing board redesigns or modifying flash drivers in their bootloaders. Flash suppliers want to be able to do whatever they feel like, so long as a thin interface layer is preserved on top....

    16. Re:Come again? by Hatta · · Score: 1

      OK, then design a slim card edge connector anyone can get to.

      --
      Give me Classic Slashdot or give me death!
    17. Re:Come again? by ajlitt · · Score: 3, Interesting

      Because:

      -there are thousands of chips out there that have a built-in SATA interface
      -BIOSes and kernels already know SATA, and developers are already used to working with it
      -MMC/SD/eMMC doesn't come close to the throughput of SATA
      -manufacturers don't like vendor lock-in, and SATA is the most popular non-embedded SSD interface

    18. Re:Come again? by doublebackslash · · Score: 1

      Part manufacturer A wants to sell a product X.
      Customer B, C, D, E, and F all use the standard that pert X conforms to.
      Profit for you, lower cost for them for using a standard part!

      As opposed to creating customer EVERYTHING for everyone.

      Or, in the vein of your signature: "Specs? That's too geeky. Just make it go." Electrical characterization and testing for custom everything isn't trivial and having standards that you know a part conforms to aids in reducing that significant engineering cost.

      So just because Best Buy can't service it does not mean that the standard is for nothing.

      --
      md5sum /boot/vmlinuz
      d41d8cd98f00b204e9800998ecf8427e /boot/vmlinuz
    19. Re:Come again? by h4rr4r · · Score: 1

      Yes some drives are that good. Even the $100 SSD I bought can soak the SATA 3.0Gbps lane I have it connected too.

      For more storage I use spinning disks in another machine, but I like my desktop to be quiet and fast.

    20. Re:Come again? by DigiShaman · · Score: 1

      Agreed! Stamping a SSD controller chip on the MB has got to be the dumbest fucking idea. That, and the use of -yet- another interface connector. What makes the SSD market so competitive and advanced are three primary components. NAND chips, the controller, and firmware. Call it the holy SSD trinity if you like, but there you go. Also, because they're already using SATA connectors, data retrieval, portability, and standardization makes the use of SSD drives available to non-PC applications. Game consoles and other DIY applications come to mind.

      No, the idea of monopolizing this immature SSD market is to grease the palm of whomever is involved by forcing a select view controller chips to be used. Focus on the I/O and interoperability, not how to control the damn things.

      --
      Life is not for the lazy.
    21. Re:Come again? by LordLimecat · · Score: 1

      A mini-sata port is absolutely tiny, and is what Apple uses in the Air (and possible the ipad). I would much rather be able to replace the thing when it inevitably dies, rather than have that extra 1x4x1 cm.

    22. Re:Come again? by LordLimecat · · Score: 2

      mini-pcie.

    23. Re:Come again? by couchslug · · Score: 1

      First they came for the IDE controllers,
              and I didn't speak out because I wasn't an IDE controller.

              Then they came for the modems,
              and I didn't speak out because I wasn't a modem.

              Then they came for the sound cards,
              and I didn't speak out because I wasn't a sound card.

              Then they came for the SSDs,
              and there was no one left to speak out for...

              Aww...fuck it. Motherboards are dirt cheap so enjoy the tradeoff.

      --
      "This post is an artistic work of fiction and falsehood. Only a fool would take anything posted here as fact."
    24. Re:Come again? by LordLimecat · · Score: 1

      MacBook Airs are flying off the shelve with RAM already soldered onto the MB

      Thats wonderful for Apple, but it certainly isnt a cost-saving measure. Im sure apple LOVES that you have to buy their ram at ridiculous $50-per-gb rates, rather than popping over to newegg and getting it for one quarter that.

      So while Im sure it is great for corporations, it is terrible for the consumer; one of the great things about PCs (and to a lesser extent, laptops) is standardized connectors that allow you to replace parts. Standardizing a system for soldering parts to the board is horrible for the consumer and prevents that.

    25. Re:Come again? by Rockoon · · Score: 1

      I also question the necessity of a faster interconnect. Are drives really sustaining those kinds of speeds?

      The first SSD maker to stress SSD performance was MTRON in late 2007. The market prior to this point was playing the "lets get bigger" game while only pushing around 20MB/sec, until MTRON's 16 GB drive turned the entire market upside down with its 100+MB/sec sustained reads. But at this point even MTRON wasn't improving write performance.

      By 2009, SSD's had been effectively saturating the SATA 2.0 link with 250MB+/sec sustained read speeds, with write speeds breaking 100+MB/sec themselves.

      SATA 2.0 wasn't even really considered universally adopted by consumers until early 2010, a year after SATA 3.0's ratification.

      Now its mid-2011, and SSD's are effectively saturating SATA 3.0 with sustained 500+MB/sec on both reads and writes, while most consumers still have only SATA 2.0 support.

      SATA is so inadequate that manufacturers are bringing back the return of the Hardcards's.

      --
      "His name was James Damore."
    26. Re:Come again? by drsmithy · · Score: 1

      Now its mid-2011, and SSD's are effectively saturating SATA 3.0 with sustained 500+MB/sec on both reads and writes, while most consumers still have only SATA 2.0 support.

      Outside of benchmarking, what are consumers doing with 500+ MB/sec of sustained transfers from a single drive ? That's a phenomenal amount of of data for a single-user PC.

    27. Re:Come again? by bmo · · Score: 1

      >Outside of benchmarking, what are consumers doing with 500+ MB/sec of sustained transfers from a single drive ? That's a phenomenal amount of of data for a single-user PC.

      You should see how snappy an ordinary PC becomes when you've got an SSD for a system and software drive.

      --
      BMO

    28. Re:Come again? by drsmithy · · Score: 1

      You should see how snappy an ordinary PC becomes when you've got an SSD for a system and software drive.

      I already know, but that's got nothing to do with bandwidth, it's all latency.

      An ordinary PC with an SSD on 6Gb SATA would be indistinguishable from an ordinary PC with an identical SSD on 100MB ATA (if such a thing existed).

    29. Re:Come again? by bmo · · Score: 1

      You're kidding, right?

      Do the math. 100MB bandwidth is a 1/6 the bandwidth of 6Gb including overhead.

      --
      BMO

    30. Re:Come again? by drsmithy · · Score: 1

      You're kidding, right?

      Nope.

      Do the math. 100MB bandwidth is a 1/6 the bandwidth of 6Gb including overhead.

      Which means nothing if you never need more than 100MB/sec of bandwidth.

      What do you think you're doing on an ordinary PC that's likely to be bandwidth limited ?

    31. Re:Come again? by bmo · · Score: 2

      You are arguing against speeding up of computers with modern technology. You are on the wrong side of history. If you wish to go that route, I suggest you might want to check out these used Hayes modems I've got because nobody could ever physically read text at over 240 chars/sec.

      Programs have grown in size over the years, in case you hadn't noticed. You may be certainly happy with your copy of PFS:Write on 8 inch floppy in your S-100 bus CP/M machine, but the rest of the world marches on. Just because *you personally* do not have any need for speed doesn't mean other people don't or shouldn't have a desire for speedy computers.

      I'll take a saturated bus at 6Gb/sec (600MB/sec (with overhead)) over ATA-6 (100MB/sec) any day.

      Your argument sounds like the yammering of an old man on his porch telling me that bias ply tires were just fine back in the 70s and should be just fine today, honestly.

      --
      BMO

    32. Re:Come again? by Anonymous Coward · · Score: 0

      This is only an issue if the motherboard isn't very expandable. As long as they have decent expansion slots, they can solder whatever the hell they like to a motherboard.

    33. Re:Come again? by drsmithy · · Score: 1

      You are arguing against speeding up of computers with modern technology.

      No, I'm questioning the suggestion that there is a genuine need for a disk interface faster than 6Gb SATA in the consumer PC space.

      I didn't say anything about there not being a need anywhere.

      Just because *you personally* do not have any need for speed doesn't mean other people don't or shouldn't have a desire for speedy computers.

      Yet the question remains. Just what are people likely to be doing on consumer PCs that is bandwidth limited ?

      I'll take a saturated bus at 6Gb/sec (600MB/sec (with overhead)) over ATA-6 (100MB/sec) any day.

      So would I, but that scenario isn't really relevant to this discussion. A sustained 600MB/sec (or even half that) is a phenomenal amount of data for a consumer PC (or, indeed, the vast majority of computers) to be reading or writing to permanent storage.

    34. Re:Come again? by bmo · · Score: 1

      >Just what are people likely to be doing on consumer PCs that is bandwidth limited ?

      It's a nonsense question.

      You may as well ask the purpose of facebook. What do people really *need* it for? What do we really need any of this for? You seem to be having trouble separating need and desire.

      --
      BMO

    35. Re:Come again? by smash · · Score: 1

      Swap.

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    36. Re:Come again? by bmo · · Score: 1

      I'm going to slap you with a fish.

      Swap on SSD is the quickest way to kill it, even with wear leveling.

      --
      BMO

    37. Re:Come again? by drsmithy · · Score: 1

      It's a nonsense question.

      Not when the premise is along the lines of "thank god, this is just what I've been waiting for".

      You may as well ask the purpose of facebook. What do people really *need* it for?

      No, not at all. If Facebook disappeared in a puff of smoke, millions of people using would notice. If the 6Gbs SATA connection in the typical desktop PC was replaced by a 100MB/sec ATA connection, hardly anyone would notice.

      You seem to be having trouble separating need and desire.

      No, I'm pretty sure I've got a good handle on "need". Which is why I say there's bugger all difference between an SSD at 100MB/sec and 600MB/sec for the typical consumer PC.

    38. Re:Come again? by drsmithy · · Score: 1

      Unlikely. Swapping tends to be short bursts of random reads and writes.

    39. Re:Come again? by Rockoon · · Score: 1

      Outside of benchmarking, what are consumers doing with 500+ MB/sec of sustained transfers from a single drive ?

      Obviously you think that consumers only deal with small datasets.

      Perhaps you are unaware that people play video games, and that the latest video games are many gigabytes (GTA 4 is 14GB), that even single maps sometimes use many gigabytes of data?

      I guess you like loading screens and progress bars. Do they turn you on?

      --
      "His name was James Damore."
    40. Re:Come again? by drsmithy · · Score: 1

      Obviously you think that consumers only deal with small datasets.

      Yep.

      Perhaps you are unaware that people play video games, and that the latest video games are many gigabytes (GTA 4 is 14GB), that even single maps sometimes use many gigabytes of data?

      Games were about the only likely candidate I could think of as well, but I'm still skeptical there's enough data being loaded from disk at any one time for bandwidth to be a genuinely limiting factor.

      I'd be very interested to see some actual numbers. How much data is being loaded at a time ? How often ? Is it exceeding 250+MB/sec reads (or writes) for extended periods of time (2-3+ seconds)?

    41. Re:Come again? by mcrbids · · Score: 1

      -MMC/SD/eMMC doesn't come close to the throughput of SATA

      Not that your other points don't have merit, but the OCZ Vertex 3 bumps up against the throughput limits of 6 Gbps SATA. Next time you might not want to make a point that's countered in the summary, unless the summary is just wrong.

      Start with ... SSDs that are fast approaching the 6Gbps ceiling imposed by the current Serial ATA specification...

      --
      I have no problem with your religion until you decide it's reason to deprive others of the truth.
    42. Re:Come again? by Anonymous Coward · · Score: 0

      No, I'm questioning the suggestion that there is a genuine need for a disk interface faster than 6Gb SATA in the consumer PC space.

      HDMI 1.3 has a theoretical bandwidth maximum of 10.2Gbps, so for a DVR that's separate from the STB that decompresses the signal, you'd need more than 6Gbps to record and playback video. Which, based on the current state of SSDs, would be about 10 minutes worth.

      Whatever the use case, uncompressed video would be the obvious candidate since the bitrate is high enough to push the boundaries of what's currently available and is likely too high to do real-time compression on anything resembling consumer hardware. Also keep in mind that hardware specs have to stay a few years ahead of the market, so they need to currently be planning for usage scenarios that are at least a couple of years away from being the norm.

    43. Re:Come again? by bemymonkey · · Score: 1

      "I already know, but that's got nothing to do with bandwidth, it's all latency."

      So all your applications are so small that they're loaded into RAM instantly just as soon as the access time (I'm assuming that's the latency you're talking about) has elapsed?

      I dunno, when I load $BigProgram, my laptop sure seems to read a lot of stuff off of the hard drive and write it into RAM... ;)

    44. Re:Come again? by ajlitt · · Score: 1

      SD, as in SD Card, SDHC, or SDXC, the fastest of which tops out at 312MB/s. Not SSD. Next time try harder to read the text you quote.

    45. Re:Come again? by ajlitt · · Score: 1

      SD, as in SD Card, SDHC, or SDXC, the fastest of which tops out at 312MB/s. Not SSD.

    46. Re:Come again? by Anonymous Coward · · Score: 0

      Well, I have been using a SSD for boot and swap on this windows machine from day one. It will probably outlast this motherboard.
      Doesn't sound any worse than having CPU, memory or GPU soldered on the motherboard.

    47. Re:Come again? by KingMotley · · Score: 1

      Haven't had a single desktop (or server for that matter) die from having swap on it's SSD yet, but then again I buy decent SSD's too. The whole wearing out the SSD thing is kinda overrated for most cases.

    48. Re:Come again? by Anonymous Coward · · Score: 0

      Yet the question remains. Just what are people likely to be doing on consumer PCs that is bandwidth limited ?

      The problem with you question is that it is retarded to begin with. The consumer does not need a computer at all to begin with, it's just convenient to have one.

      I don't need my computer to boot in less than 10s, but I think it is worth to spend money on a SSD to have it that way.
      No-one have a need for a sustained 600MB/s but if you are going to move that 100GB picture archive or whatever from one disk to another it is pretty convenient to have that kind of bandwidth.

      Example of other things people doesn't really need but think is nice to have: Automobiles, Electricity, Telephones, Nail polish, Mirrors, Books, Underwear.

    49. Re:Come again? by KingMotley · · Score: 1

      I think more people would notice than you think. A single hard drive from ~5 years ago could easily saturate that 100MB/sec ATA connection.

    50. Re:Come again? by brentrad · · Score: 1

      For one thing, I don't think the article mentioned consumer drives specifically. These standards are just as applicable for servers - and for servers, the faster the drive, the better. Arguing that these SSDs are "fast enough" for servers is ludicrous - you can never have too much speed in your servers.

      But I'll give you an example of large datasets that a consumer (me) would use these types of speed for, on a daily basis: I download large HD video files from newsgroups, and the program first has to assemble the posts into RAR files, then check the files for corruption, then assemble and unpack them. These files can be 10's of GB, and all the above processes are very I/O bound. If I could run them on an SSD, they would complete in a fraction of the time, which means I could download even more.

      Installing an OS is also very dependent on the speed of your drive. If you can reinstall your OS in 12 minutes (that's how long it took with my SSD to install Windows 7) vs 45 minutes with a hard drive, that's a big gain. If you're a member of your IT department cloning a bunch of computers, that's a huge increase in your productivity.

      Does my mother need such a fast SSD drive in her computer? No, of course not, she'd be fine with a slow SSD like you describe. But not every consumer wants "good enough" in their computers - that's why we have a range of options and enthusiast parts. I, for one, will choose the faster drive. I won't always use the increased speed, but when I need it I will sure appreciate it.

    51. Re:Come again? by Anonymous Coward · · Score: 1

      mini-pcie.

      And there is also SSD drives designed for this form factor. Seriously pissed off when I found out my new Lenovo didn't support it, but the next model up did. Could have had a fast SSD and space for a standard 2.5" drive - in a laptop, that is perfection.

    52. Re:Come again? by drsmithy · · Score: 1

      Whatever the use case, uncompressed video would be the obvious candidate since the bitrate is high enough to push the boundaries of what's currently available and is likely too high to do real-time compression on anything resembling consumer hardware.

      I can guarantee you compression is a more viable solution than disk bandwidth. Particularly with the powerful GPUs systems come with these days.

      I can't see many, if any, home users dealing with uncompressed video. The space requirements, even before the performance requirements, are prohibitive.

    53. Re:Come again? by drsmithy · · Score: 1

      The problem with you question is that it is retarded to begin with. The consumer does not need a computer at all to begin with, it's just convenient to have one.

      Yes, if you want to go all Reducto ad Absurdum, there is no point answering the question.

      In the real world, of course, there is a whole spectrum of grey between what you need to survive, and what you might dream about having.

    54. Re:Come again? by drsmithy · · Score: 1

      I think more people would notice than you think. A single hard drive from ~5 years ago could easily saturate that 100MB/sec ATA connection.

      Er, yes, in highly specific conditions.

      Even today, you rarely see that sort of sustained performance from drives, even if they can sustain 150M/sec in benchmarks.

      Having done a lot of performance profiling in my time, I think I've got a reasonable handle on disk bottlenecks. Bandwidth is rarely one of them.

    55. Re:Come again? by bmo · · Score: 1

      But obviously you haven't done performance testing on SSDs.

      SSDs are not limited to the rate of the spinning physical media. They can be read much faster. When a hard disk runs out of its cache, the SSD is still delivering data *at the maximum rate.* There is no dropoff as you exhaust "cache" because there is no cache to exhaust

      That's what you're not getting. It's not just seek time difference, an SSD behaves as if it has a cache the same size as the disk itself.

      --
      BMO

    56. Re:Come again? by KingMotley · · Score: 1

      I hit it all the time on both my home machine and work machine. Copying large files (video), doing subversion updates, or dropbox scans/indexes, rebooting, etc etc

    57. Re:Come again? by Dog-Cow · · Score: 1

      Until static storage (HDD, SSD, whatever) is as fast to access as main system memory, it's not fast enough. I am not sure why this is hard to understand. The purpose of a tool is to make operations more efficient. The less time taken, the more (time) efficient the process. If you're not trading anything for that efficiency gain, why would you not take it?

    58. Re:Come again? by Dog-Cow · · Score: 1

      There is a limit to how small/thin a device can be if soldering is not used. You may not care to have devices that break this limit, but millions do.

    59. Re:Come again? by Dog-Cow · · Score: 1

      It's done so that the housing can be thinner and lighter. You may not care for the tradeoff, but millions do, and they matter to Apple and you don't.

    60. Re:Come again? by jedidiah · · Score: 1

      No. It's not a nonsense question.

      Until you identify the need, or the use case it's hard to create a meaningful design.

      Rambling on about faster connectors on tablets with no storage to speak of in the first place is pretty absurd.

      OTOH, I would love to be able to move large amounts of stuff around my non-portable storage a bit faster. Although current affordable consumer storage devices need to catch up to current interconnects first before the cable becomes the bottleneck.

      People are simply fixating on the wrong bottleneck on the wrong device.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    61. Re:Come again? by jedidiah · · Score: 1

      No you aren't. Really you aren't.

      It doesn't sound like you have enough stuff for the slowness of current devices to be that painful.

      Also, if you are going to bring "drop box" into this then the local storage tech doesn't even matter at all. You're bottlenecked by the network then.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    62. Re:Come again? by HappyPsycho · · Score: 1

      Anyone who has more than 1 hard drive will notice.

      Having to ensure my hard drive and CD burner were on different cable to ensure that a consistent stream could be maintained sounds like a need for a higher speed link, note I said cd burner not dvd.

    63. Re:Come again? by jedidiah · · Score: 1

      You should not be swapping on a modern machine.

      This is not 1995. Memory is cheap enough that your machine should have enough of it even with today's bloated operating systems.

      If your OS benefits greatly from simply adding an SSD, then it's probably broken.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    64. Re:Come again? by drsmithy · · Score: 1

      But obviously you haven't done performance testing on SSDs.

      Yes, I have.

      That's what you're not getting. It's not just seek time difference, an SSD behaves as if it has a cache the same size as the disk itself.

      I get it just fine. The point you seem to be missing is if there's nothing on the other end actually demanding all that data, then the fact so much of it can be delivered is irrelevant. The other part of that point is, that outside of benchmarks, there are very few (if any) tasks on the ordinary PC that demand data transfers that fast, for long enough to matter.

    65. Re:Come again? by drsmithy · · Score: 1

      I hit it all the time on both my home machine and work machine. Copying large files (video), doing subversion updates, or dropbox scans/indexes, rebooting, etc etc

      I sincerely doubt you're hitting the limits of the SATA interface doing that. Especially since a lot of the operations you mentioned are random access, you'll never even get close to streaming performance with them.

      You might be hitting the limits of how fast the mechanical disk can deliver data to the interface. It's highly unlike you're hitting any limits of the interface.

    66. Re:Come again? by drsmithy · · Score: 1

      Until static storage (HDD, SSD, whatever) is as fast to access as main system memory, it's not fast enough. I am not sure why this is hard to understand.

      It's not at all hard for me to understand. I'm not even disagreeing.

      The purpose of a tool is to make operations more efficient. The less time taken, the more (time) efficient the process. If you're not trading anything for that efficiency gain, why would you not take it?

      My point is it's unlikely there's any operations on "the ordinary PC" that are coming anywhere close to the limits imposed by SATA3 (or SATA1, for that matter), so there's not going to be any gain from making it faster.

    67. Re:Come again? by drsmithy · · Score: 1

      Anyone who has more than 1 hard drive will notice.

      Won't matter. SATA bandwidth limits are per device.

      Having to ensure my hard drive and CD burner were on different cable to ensure that a consistent stream could be maintained sounds like a need for a higher speed link, note I said cd burner not dvd.

      It's not the 1990s, and we'not dealing with ATA any more. Even if we were, the problem wasn't that the link was fast enough, it was that the protocol didn't allow two devices on the same cable to be active at the same time. A SCSI interface from the same era (and of slower performance) than ATA100 could daisychain half a dozen CD burners and still work.

    68. Re:Come again? by drsmithy · · Score: 1

      So all your applications are so small that they're loaded into RAM instantly just as soon as the access time (I'm assuming that's the latency you're talking about) has elapsed?

      They're certainly small enough that the difference between 600MB/sec (or even 100MB/sec) and anything faster is irrelevant.

      I dunno, when I load $BigProgram, my laptop sure seems to read a lot of stuff off of the hard drive and write it into RAM... ;)

      Well, get an SSD drive and you'll see a big difference.

      Newer versions of Windows actually make this pretty easy to examine. Just fire up the "Resource Monitor" and do some stuff. If the graph for "Disk IO" isn't frequently bumping up against 150MB/sec, you wouldn't even be slowed down by SATA1.

    69. Re:Come again? by KingMotley · · Score: 1

      When you first start your machine, dropbox has to index all the files on your mirror. I assume it's doing an internal hash (of all 50GB worth of files), and then once it's complete, it then compares those hashes to the values on the server, so yes, it is very local I/O intensive.

      Would a Youtube video showing a realtime throughput be enough proof?

    70. Re:Come again? by bmo · · Score: 1

      >The point you seem to be missing is if there's nothing on the other end actually demanding all that data,

      Let me introduce you to my music collection when I transfer it from one drive to another.

      You are arguing, continually, from your own little point of view and applying it to everyone, saying that you are a typical user. What utter nonsense, and what hubris. You are also arguing from incredulity.

      I'm done here. You will not concede the point that people are different from you and have different needs or desires.

      --
      BMO

    71. Re:Come again? by KingMotley · · Score: 1

      Here's some hard numbers of typical things I do:

      Running Videoscripts Metadata Batcher to retag some videos: 166MB/s
      Copying video files from my C: to B: drive: 170-220MB/s

      Both of these would exceed the 100MB/s ATA link mentioned, and NEITHER of these are using my SSDs which are many times faster.

    72. Re:Come again? by drsmithy · · Score: 1

      Both of these would exceed the 100MB/s ATA link mentioned, and NEITHER of these are using my SSDs which are many times faster.

      Your second example is clearly using a RAID array, so the first probably is as well.

      Not only is this far from an "ordinary PC", but only at the very upper end (220MB/sec) *might* it really become limiting, because that 100MB/sec is per device.

      Not to mention, it's still far away from the 300MB/sec of SATA2, and not within a bull's roar of the 600MB/sec of SATA3.

    73. Re:Come again? by drsmithy · · Score: 1

      Let me introduce you to my music collection when I transfer it from one drive to another.

      Right. What are you copying *to* ?

      You are arguing, continually, from your own little point of view and applying it to everyone, saying that you are a typical user. What utter nonsense, and what hubris. You are also arguing from incredulity.

      Actually, no, I haven't said anything about my needs, wants, behaviour, or anything else. I've just made the point that 600MB/sec *per device* is a phenomenal amount of data, and it's very hard to see anything in the "ordinary PC" space that would benefit in the slightest from greater speed.

      Heck, even most servers would rarely see sustained disk performance past 100-200MB/sec.

    74. Re:Come again? by drsmithy · · Score: 1

      For one thing, I don't think the article mentioned consumer drives specifically. These standards are just as applicable for servers - and for servers, the faster the drive, the better. Arguing that these SSDs are "fast enough" for servers is ludicrous - you can never have too much speed in your servers.

      The article may not have specified consumer PCs, but the person I responded to did explicitly say "ordinary PCs". Though even most servers would not benefit from >600MB/sec per disk device. Heck, most of them wouldn't see any benefit from >150MB/sec, since individual drives that can exceed this outside of corner cases are still relatively uncommon, and there's generally not anything capable of *consuming* the data any faster.

      But I'll give you an example of large datasets that a consumer (me) would use these types of speed for, on a daily basis: I download large HD video files from newsgroups, and the program first has to assemble the posts into RAR files, then check the files for corruption, then assemble and unpack them. These files can be 10's of GB, and all the above processes are very I/O bound. If I could run them on an SSD, they would complete in a fraction of the time, which means I could download even more.

      Somehow I doubt the speed at which you can reassemble your movies is limiting how fast you can download. Unless you're mainlined into your Usenet server at a gigabit.

      Also, your ability to reassemble - even with an SSD - is going to be limited by the write performance, which still isn't going to exceed 6GB/sec for any long period of time.

      Installing an OS is also very dependent on the speed of your drive. If you can reinstall your OS in 12 minutes (that's how long it took with my SSD to install Windows 7) vs 45 minutes with a hard drive, that's a big gain. If you're a member of your IT department cloning a bunch of computers, that's a huge increase in your productivity.

      Again, the limiting aspect there is the internal performance of the disk device itself, not its connection to the rest of the system.

      Does my mother need such a fast SSD drive in her computer? No, of course not, she'd be fine with a slow SSD like you describe. But not every consumer wants "good enough" in their computers - that's why we have a range of options and enthusiast parts. I, for one, will choose the faster drive. I won't always use the increased speed, but when I need it I will sure appreciate it.

      I'd be pretty confident in predicting even with the fastest SSDs today or within the next few years, you wouldn't be meaningfully constrained by a 600MB/sec per device limit for anything close to normal usage.

    75. Re:Come again? by brentrad · · Score: 1

      ...and 640k ought to be good enough for anybody.

    76. Re:Come again? by bemymonkey · · Score: 1

      So when you load a 500MB application from disk, whether it's loaded in under 1 second or in 5 seconds is irrelevant?

    77. Re:Come again? by drsmithy · · Score: 1

      So when you load a 500MB application from disk, whether it's loaded in under 1 second or in 5 seconds is irrelevant?

      I can't think of any applications I have that want to load up 500MB of data at startup.

    78. Re:Come again? by smash · · Score: 1

      Swap is still useful. Why the fuck keep programs/DLLs/libraries in RAM that are not actively running when that RAM could be better used for disk cache for programs that are actually active?

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    79. Re:Come again? by adisakp · · Score: 1

      It's done so that the housing can be thinner and lighter. You may not care for the tradeoff, but millions do, and they matter to Apple and you don't.

      Exactly, the sockets and extra daughterboards for expandable memory take up some weight and space. If you are trying to be on the bleeding edge of thin and light and still be reasonably priced (the MacBook Air is Apple's cheapest laptop at $999), you have to make some tradeoffs and internal expandability is the easiest one to make -- especially since 99% of notebook owners never change a thing inside their computer.

  2. Yeah... by Anonymous Coward · · Score: 0

    That was about my response as well.

  3. Hmmm by Anonymous Coward · · Score: 1

    I can't imagine anything i'd want my tablet to have 16Gbps disk speeds for..

    That just seems like a stupid waste for a tablet or other small form device.

    1. Re:Hmmm by jedidiah · · Score: 1

      Paradoxically, tablets don't have enough storage for storage speeds of that magnitude to be terribly compelling.

      SSDs are relatively puny and storage on tablets even more so.

      You end up with a device capable of saturating a fiber connection being connected to 3G, or bluetooth, or USB.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    2. Re:Hmmm by wagnerrp · · Score: 3, Insightful

      At those kinds of speeds, you start talking about a system that goes from completely off to completely on in a second. When you want to hibernate, you dump everything in memory over to the disk. When you turn back on, you take a moment to find the disk, and pull the entire memory image back over. There is no boot, there is no shutdown. You only need enough memory to handle the actual in-use programs, and anything else could be painlessly paged out, meaning you never have to close programs.

      It's an order of magnitude slower than RAM, but an order of magnitude faster than hard disks. Right smack in the middle in order to offer all sorts of cool little tricks.

    3. Re:Hmmm by fuzzyfuzzyfungus · · Score: 1

      It might make running up against the pathetically tiny supply of RAM a bit less painful...

    4. Re:Hmmm by h4rr4r · · Score: 1

      I would like mine to turn on instantly and not totally stall out when I run out of ram or have to hit the disk for some reason.

      Get a cheap SSD and boot from that. I used to leave my desktop on all the time, now it boots faster than it used to recover from sleep/standby.

    5. Re:Hmmm by jedidiah · · Score: 1

      You don't even need to get to those kind of speeds to start treating your "storage" as bubble memory.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    6. Re:Hmmm by 0123456 · · Score: 1

      You don't even need to get to those kind of speeds to start treating your "storage" as bubble memory.

      Because repeatedly writing to media with a limited number of write cycles is such a good idea.

    7. Re:Hmmm by h4rr4r · · Score: 1

      When was the last time you killed an SSD that way?
      My laptop has one, and so far over the last 3 years I have seen no such issue. It even had a swap partition for quite a while.

    8. Re:Hmmm by 0123456 · · Score: 1

      When was the last time you killed an SSD that way?

      Me, never, because I've deliberately configured the system to ensure it minimises writes to the SSD.

      Some guys who use their SSD for compilation were saying recently that they have to replace them at least once a year when they hit the write limits and the SSD dies (which makes sense for them as the programmer time saved more than pays for a new SSD every year). You'd probably burn through them even faster than that if you were using them for fast swap space on a machine with limited RAM.

      And then, of course, since it's soldered onto the motherboard it's time to buy a new tablet/phone/whatever.

    9. Re:Hmmm by orange47 · · Score: 1

      hibernation is bad idea because of memory leaks and similar.

    10. Re:Hmmm by wagnerrp · · Score: 1

      You sound like a Windows 98 user.

    11. Re:Hmmm by orange47 · · Score: 1

      and you sound like a troll.. its not a good idea to write several Gb to a flash storage on every shutdown just to have slightly faster boot.
      perhaps boot wouldn't be faster at all, given that it doesn't matter much to SSD whether its sequential or random read.

    12. Re:Hmmm by Neil+Boekend · · Score: 1

      Don't load it with an AI, it would get depressed. "Here I am, with an internal connection of 16GB/s, and they ask me to send a file over bluetooth, with only 24Mb/s"

      --
      Well, I might have a way, but it only works on a semi spherical planet in a vacuum.
  4. Oh boy! by Anonymous Coward · · Score: 0

    Now when the SSD goes out(and they do fail, sometimes faster than harddrives) the whole computer is trash without doing reflow on the motherboard. This idea was clearly made by OEMs who prefer you buy another computer than fix the one you have

    1. Re:Oh boy! by X0563511 · · Score: 1

      That's why it's tested before it's soldered on.

      --
      For large sets, this will be our guide even unto death, for the LORD will work for each type of data it is applied to...
    2. Re:Oh boy! by LordLimecat · · Score: 1

      It will eventually fail without much recourse for 98% of consumers.

    3. Re:Oh boy! by X0563511 · · Score: 2

      So, you mean that 2% will continue to work until the end of time?

      --
      For large sets, this will be our guide even unto death, for the LORD will work for each type of data it is applied to...
  5. Sounds like OCZ's IBIS by denis-The-menace · · Score: 2

    Sounds like OCZ's IBIS just got standardized or copied.

    http://www.ocztechnology.com/ocz-ibis-3-5-high-speed-data-link-ssd.html/

    --
    Obama's legacy: (N)othing (S)ecure (A)nywhere and (T)error (S)imulation (A)dministration
    1. Re:Sounds like OCZ's IBIS by fuzzyfuzzyfungus · · Score: 1

      OCZ was hardly the first to do direct flash-PCIe storage(though some of their earlier products were just a disk controller and SSD on the same card, they have native ones now). They do have the advantage of being one of the vendors of PCIe-SSDs whose prices are remotely accessible, and who are available through enthusiast channels.

      Most of the other players are basically in the business of making people's Big Serious Expensive databases run faster, and their prices and "if you are serious, please call our sales department" distribution style. You probably don't want to know what one of these or these cost, and you won't find them at newegg...

    2. Re:Sounds like OCZ's IBIS by Anonymous Coward · · Score: 0

      It's just $15,000. What are you whinging about?

    3. Re:Sounds like OCZ's IBIS by 0123456 · · Score: 1

      It's just $15,000. What are you whinging about?

      My first SSD cost $50,000 and that was in 1990 money.

  6. read a little closer by Anonymous Coward · · Score: 0

    SSD controller, not the drive itself.

    1. Re:read a little closer by Osgeld · · Score: 1

      the controller just does all the work specific to the flash ram installed

    2. Re:read a little closer by unixisc · · Score: 2

      If the SSD drive is on the PCI-X card itself (which it'd have to be to attain that 16Gb/s speed), then why not put the SATA controller chip on that, instead of the motherboard? This is assuming that the motherboard is not going to support SATA HDD's as well that are separately connected to it. But if a SSD is connected, have a way for the motherboard controller to be disabled and unused, or used only for a connected HDD, would make more sense. I disagree w/ the others who've said that the entire SSD is in the chip. No, given the densities you have for NAND flash, you need several of those to get to the densities desired. The controller has to be connected to the NAND flashes in order to do things like ECC and file system management.

  7. Dig a little deeper, Watson by Anonymous Coward · · Score: 0

    It IS the entire SSD in a single solderable package.

  8. Funny how things come full circle by smead · · Score: 4, Interesting

    Hard drives in PCs start out with a proprietary interface by Segate that becomes a defacto standard. It needs an interface controller to tell the drive what to do. That controller sits on the ISA bus. Speeds increase, drives become bigger, they move the controller onto the hard drive. The ISA bus still connects to the controller, and the controller still tells the drive what to do, it's just that we now call the connection between the motherboard and the controller the IDE (integrated drive electronics) bus, but it's still the ISA bus. Speeds increase, now we increase the speeds of the IDE bus and add features, it slowly moves away from the ISA bus as the IDE controllers get more complicated. Speeds increase and having that bus as a parallel interface doesn't cut it, so we invent SATA. A SATA controller sits on the PCI bus and tells the drive's controller what to do. Speeds increase and now we're back to directly connecting the hard drive to the PCI (now PCI-E, but same parallel to serial transition) bus. -- Full Circle.

    1. Re:Funny how things come full circle by Anonymous Coward · · Score: 0

      "Full Circle."

      Let us not forget the whole reason that there is a full circle is the advent cheap (relatively speaking) chip-based long term storage. So the only reason we are doing this is - because chip-based tech has finally become possible.

      The whole argument rests on the advent of key advances being in place at a given time and when you take the measurement.

  9. NVM Express already released. by Anonymous Coward · · Score: 0

    The NVM Express spec is already released. This is targeted for PCIe SSD's.

    http://www.nvmexpress.org/

  10. The potential of such SSDs by unixisc · · Score: 1

    I agree w/ you on stamping a controller chip on the MB. But the use of the interface connector - if an SSD is orders of magnitude faster than an HDD, and there's not much else on the PCI-X bus, then why not? Only case where I can think of a justification is notebooks & tablets, where you don't have extra SATA slots, and the device could use a lower power consumption as well. Another thing worth noting - SSD comes into the market at almost the same time that the industry is moving to 64-bit. That offers the potential for major paradigm changes - one can have a memory map where part of the region is dedicated to DRAM, another dedicated to SSD, another to I/O devices and so on, w/o worrying about ever running short. If such a thing is attained, then it makes sense to try and maximize the speed of SSDs (as long as one doesn't reduce their long term endurance or data retention in the process. For game consoles and other such apps, an SSD w/ an USB interface would be more appropriate - just connect an external drive to a console w/o mucking w/ SATA buses or what else, and just have the thing work. Use SATA purely for internal purposes. It's not always about monopolizing - it's just a case of making good use of existing standards.

  11. SSD don't do random read/writes by unixisc · · Score: 1

    Since the internals of an SSD are NAND flash, it will not be a random read. NAND flash devices read or write data in pages, and the sector/block size defines how much you read or write. But you never read a byte, or word, or quad-word @ a time: you read a complete page. If random reads or programming was needed, one would use a NOR flash for that purpose. As a NOR flash is what is used when the system boots, and contains all the configuration info, it's probably worthwhile to expand its density some to capture whatever is needed when PC needs to go instant on. The HDD ain't where the system goes to find such things, and all systems that have instant on don't come w/ an SSD. NOR flash - the 4Mb flash that's typically on motherboards - is the key when it comes to rebooting.

    1. Re:SSD don't do random read/writes by orange47 · · Score: 1

      then why not define logical block size to be as large as NAND page? or defragment the thing. if that NAND page is rather large, OS makers should refrain from using multitude of small files.