Slashdot Mirror


DSI Delivers up to 3GB/s with Solid State Disk

olivesaregross writes "'Running at what the company says is 250 times the speed of conventional hard drives, it won't come cheap, but it will be fast. It uses DRAM memory to store data instead of spinning platter hard drives, giving an access time of just 20 microseconds.' It still does use platter-based drives but it's a cool idea anyway. Techworld has another story on it."

21 of 214 comments (clear)

  1. Not cheap, but fast by Anonymous Coward · · Score: 4, Insightful

    then it must be good. cheap, fast, good, pick any two.

  2. Interesting Uses Possible by Pavan_Gupta · · Score: 4, Interesting

    I can imagine this kind of technology being really applicable in situations where large databases are in use -- but potentially, slightly cheaper then just keeping the entire database in ram. I think it would be interesting to use, but a bit more interesting to play with.

    1. Re:Interesting Uses Possible by Neon+Spiral+Injector · · Score: 4, Interesting

      But the entire DB is in RAM. But it isn't local to the processor, it is on the other side of a FC bus.

      When this makes sense, is when you do want your entire DB in RAM, but don't want to put the required amount of RAM in each node of a cluster. So you just attach each node to this SAN.

  3. Looks like those companies... by Anonymous Coward · · Score: 5, Funny

    Are already starting the hard work on getting the components together to meet the minimum specs for running Longhorn. Huzzah!

  4. SSD is a niche technology at best by Anonymous Coward · · Score: 4, Informative

    Let's see:

    Imperial folded, Platypus folded, Solid Data is barely hanging on and Texas Memory survives on defense contracts.

    SSD is a great technology, yes.
    SSD makes commercial sense, no.

    How many more VCs can be fooled into investing into SSD startups?

  5. RELIABILITY!!! by NineNine · · Score: 4, Insightful

    Screw speed. At least for me, that's not an issue. I want a r-e-l-i-a-b-l-e hard drive. I've tried all the brands, but they all come down to this: You have moving parts. It's going to break, eventually. The bearings will go. The head will hit a platter, etc. Personally, I've been saying for years that a solid state hard drive will be the next big boost in PC technology, and I think this is the beginning.

    I don't understand why the company isn't touting reliability. If I have a slow hard drive... eh. No biggie. I wait an extra second. If I have a hard drive crash, that's potentially days of lost work and business, even more if a backup failed recently. I'll be buying these just as soon as I can afford them. With these drives in place, the next reliaibility bottleneck are the stupid little cooling fans failing. Electronics (printed circuit boards, chips) rarely fail on their own. It's almost something with moving parts (like a fan) that leads to their death.

    To me, this is the most exciting advance in computing since Ethernet.

    1. Re:RELIABILITY!!! by neurojab · · Score: 4, Insightful

      >I don't understand why the company isn't touting reliability.

      Traditionally, DRAM-based storage units are LESS reliable than hard disks. Why? Power loss. Yes, you can always create a massive UPS, but to be really considered "stable storage", you need to be able to store data without power for years on end. The advantage here is that the unit writes data to its hard disks, giving you some assurance that you won't lose your data even then. That turns the RAM in the unit into a giant cache. This helps with read operations, but does nothing for writes. Besides, if the hard drives fail in this unit, the unit still fails. That makes it no more reliable than RAID.

      >To me, this is the most exciting advance in computing since Ethernet.

      This concept has been around since before Ethernet. The concept of storage in solid-state isn't new. Even using RAM for a hard disk isn't new. Ever run VDISK in an expanded-memory DOS system? This concept was available in the higher-end comptuer world far before that.

  6. Google Cache by DeathToBill · · Score: 4, Informative

    Hmmm, one of the fastest slashdottings in recent history, methinks. Google cache is here.

    --
    Slashdot - News for Nerds, Stuff that Matters, in ISO-8859-1 Has just realised that beta makes this signature redundant
  7. Slashvertisement? by stratjakt · · Score: 4, Interesting

    It still does use platter-based drives but it's a cool idea anyway.

    From the article, I gather these are merely SAN boxes with up to 64GB of DRAM, fiber channel output, and 3 hot-swappable hard drives that act as backup.

    Has a record been broken? Has anything special happened? Sure this is high-end stuff, but it doesnt seem new or particularly exciting.

    --
    I don't need no instructions to know how to rock!!!!
  8. Re:Pricing by corbettw · · Score: 4, Funny

    I may be wrong... but I cannot RTFA.. the site is /.tted...

    Apparently, increasing memory access times wasn't the panacea they thought it would be.

    --
    God invented whiskey so the Irish would not rule the world.
  9. DSI Delivers up to 3GB/s... by ForestGrump · · Score: 4, Funny

    damn and I thought I could get a 3GB/s DSL line...

    --
    Is it true that more people vote for the winner of American Idol, than vote for the president? -Ali G.
  10. Re:Obligatory first post! by morcheeba · · Score: 4, Funny

    At 3 GB/sec, you can get 3000x3000 uncompressed video at 100 Hz refresh rate, with 1600 channels of uncompressed 96kHz 32-bit sound. Or compress it and get 35000x35000 resolution.

    Pure bliss. But, at that rate, the largest drive (64GB) would only last 21 seconds.

  11. Consumer edition by swb · · Score: 5, Interesting

    How many 1G sticks of RAM could you put into your standard Firewire/USB2 enclosure? Why couldn't someone make a USB2/Firewire/SCSI enclosure that the host system saw as a mass storage device but was actually just a smaller version of the above? It might be really useful for some DB applications, video editing, etc.

    I can't imagine that an enclosure of that type would run more than $500, plus the cost of the RAM that went into it. It might not be consumer cost effective, but it could be worthwhile at the prosumer or low end, where the RAM disks shown on /. are almost never affordable but by the richest organizations.

  12. Re:Pricing by eht · · Score: 4, Informative

    RAID has a couple of semi accepted meanings,

    The I can mean either Inexpensive or Independent
    and the D can be either Drives, Disks, or Devices.

    However it's always a Redundant Array, which of course makes RAID 0 not RAID, but just a good way to lose even more data faster(as any drive/disk/device failing on RAID 0 takes down all your info).

  13. Used interface? by Karamchand · · Score: 4, Insightful

    What interface are they using? Even the fastest SCSI can't provide 3GB/s!

    1. Re:Used interface? by chabotc · · Score: 4, Informative

      Oh and reading in the article it actually tells you whats used to obtain the 3gb/sec

      "with two to eight Fibre Channel ports that can push out 250,000 IOPS - up to 3Gbit/s"

  14. Now that's a drive... by theendlessnow · · Score: 4, Funny

    ...for Longhorn!!

  15. Solid State Discs have been around for a while by Burdell · · Score: 4, Interesting
    I've got a mail server that uses an SSD for the mail queue filesystem. It is great for that because the random I/O transactions per second rating is 10,000 (vs. a typical hard drive thrashing hard at 150 tps).

    The SSD we have is a Nitro!Xe from Curtis, Inc.. It looks like a standard 3.5" wide 1" high Ultra2 SCSI drive with an 80 pin SCA connector. We have a 2G model with a 2.5" notebook drive for backup (it has a battery to dump RAM to disk on power off) and it greatly improved the performance of our mail server (high performance mail queue is all about I/O TPS).

  16. Wafer-scale integration? by steveha · · Score: 4, Interesting

    Whatever happened to wafer-scale integration?

    I read an article about this years ago. The idea was something like this:

    Memory chips are made on wafers. They are made side-by-side, then sliced apart, then each tested and mounted in a package. (Then eventually mounted on a little circuit board, and thence into our hands to install into our computers.)

    The idea was to make a wafer of memory chips, but not to just have them side-by-side; actually have traces connecting one chip to another. Then use the whole wafer as a RAM unit. You would need to test and find any defective RAM chips in the wafer, then cut a trace (or burn out a fuse, or whatever) to disconnect them from the rest of the wafer. (Not too different from bad-block management on a hard drive, really.) Finally you could make a stack of these wafers in a box, and sell it as a disk drive.

    This should be much cheaper than current RAM-based disk drives. It would be slower (the traces connecting the chips would be slower than a direct memory bus to each chip) but still way faster than a drive with moving parts.

    My understanding is that wafer-scale integration isn't very interesting for most applications, but for the specific niche of RAM-based storage units it seemed promising. Clearly I'm wrong since it didn't happen. Anyone know why?

    steveha

    --
    lf(1): it's like ls(1) but sorts filenames by extension, tersely
  17. Re:Performance and Cost by MrChuck · · Score: 4, Informative
    1. PCI @ 150MHz * 64bits is still 1200MB/s (that handwaves away any sort of overhead a bus might impose). So how are they getting 3036MB/s into a machine? Certainly not FCAL or SCSI. magic?
    2. When offered SSDs to work with for a project that needed lots of writes/reads of short lived files, I found them expensive, but good. I'm sorry to hear that Platypus ate it, I liked them and the idea of RAM on a PCI card - and they had linux and freebsd drivers. But SSDs COST a BUTTLOAD.
    3. I've used these guys' disks for years because they have tested to be as fast as SSDs. But with a half terrabyte behind them.

      With a battery backed cache of mirrored RAM, we found that for quick read/write stuff, the disks never got hit. If the data stayed, they ended up on the drives. If power was lost, the battery kept the cache alive for well over a day (I got bored and it met the "30 minutes" criteria we were looking at).

      The cache isn't huge (512? 256MB?) but it never filled. Basic elevator algorithms (we all did CS classes, right?) let the RAID side take data out of the cache in DISK order and write it out.

      And, not being Computer Vendor RAID, we found that it was fast and not expensive (given professional RAID). 15KRPM disks and dual controllers and dual PS and all that. Not for home use, but certainly for pro use. Oh and it gives great stats. Find stripe usage and cache hits on a Sun T3 that performs at half the speed for a good bit more money.

    I don't work for them, I just like their stuff. They're a small(ish) company that just does raid with lots of Wall St and corporate clients.
  18. The RAM itself is one thing... by Kjella · · Score: 4, Interesting

    ...that you can actually do in software (get a PowerMac or AMD-64, load it up with 8-12GB of ram). What usually cost money, is to have some sort of flush-to-disk feature.

    I'd love a SSD at least big enough to boot from, to combine with some other fanless stuff to create a 100% fanless, no moving parts PC (except from burner, which is silent when not in use). That + GbLan (to copy everything in from fanless machine, no damn spinning CD/DVD) using a direct crossover cable to a file server, preferably in a sound-isolated galaxy far, far away.

    That is my dream for my next setup. I've looked at doing the same simply dragging DVI + USB cables + external burner at machine, but it's not that great. A network cable can go so much longer...

    Kjella

    --
    Live today, because you never know what tomorrow brings