Forgot your password?
typodupeerror
Data Storage

Crucial Launches MX100 SSD At Well Under 50 Cents Per GiB 107

Posted by Unknown Lamer
from the cheaper-better-faster dept.
MojoKid (1002251) writes "Crucial has been on a tear as of late. In the last few weeks alone, the company has released a couple of new series of solid state drives, one targeting the enthusiast segment (the M550) and the other targeting data centers (the M500DC). Today, Crucial is at it again with the launch of the brand new MX100 series. The Crucial MX100 series of solid state drives is somewhat similar to the M550 in that they both use the same Marvell controller. The MX100, however, is outfitted with more affordable 16nm NAND flash, and as such, the drives are priced aggressively at about .43 per GiB. However, these MX100 series of drives are still rated for 550MB/s sequential reads with 500MB/s (512GB), 330MB/s (256GB), or 150MB/s (128GB) and random read and write IOPS of 90K – 80K and 85K – 40K, respectively. The drives carry a 3-year warranty and are rated for 72TB total bytes written (TBW), which equates to 40GB written per day for 5 years. Performance-wise, these new lower cost SSDs, are on par with some of the fastest SSDs currently on the market but starting at $79.99 for the 128GB drive, they're relatively rather cheap."
This discussion has been archived. No new comments can be posted.

Crucial Launches MX100 SSD At Well Under 50 Cents Per GiB

Comments Filter:
  • Ye Gods, an Ad (Score:5, Insightful)

    by debatecoach (2883733) on Monday June 02, 2014 @06:03PM (#47150607)
    Is this supposed to be informative, or an ad? Has Crucial purchased a stake in slashdot?
    • And bonus points to anyone who can correctly point out the literary reference in my post...
      • The Music Man...
        • Ok I'll give it to just think it. I was thinking of "The Name of the Wind" by Patrick Rothfuss but apparently the colloquialism exists beyond my experience. A great read by the way. Which is ironic, as I criticized the ad and included an ad for my favorite series...
      • Re: (Score:3, Funny)

        by bjwest (14070)
        Dracula. Stake is referenced quite often in that story.
    • Do you think the editors crack is free?

      Might be: 'Free' as in 'Free beer for speech'.

    • Re:Ye Gods, an Ad (Score:5, Insightful)

      by mcrbids (148650) on Monday June 02, 2014 @06:47PM (#47150895) Journal

      Perhaps it's an ad, but it's one that interests me. I come here to find out the latest developments in tech, and the continuing advances of SSDs is something I find interesting.

      HDDs have become so huge that the biggest problem isn't storage capacity, or even bandwidth: it's IOPS. It's pretty lame that I can store literally many millions of documents in a hard disk cluster that can only delivery a few hundred IOPS per second. Do the math. It takes forever to get your data out, especially if they are small documents!

      SSDs don't have this problem. 50,000 IOPS is "no big deal" for an SSD, meaning that even if you have 40 million tiny 10k documents, you can still saturate your 6 Gbit SATA interface with sweet, sweet data.

      We switched our DB servers to SSDs and saw over 90% reduction in average query latency. Next up is our file stores, which use ZFS. Our next step is an SSD cache for ZFS, and then as prices continue to tumble, we'll switch to all SSDs everywhere.

      • We switched our DB servers to SSDs and saw over 90% reduction in average query latency. Next up is our file stores, which use ZFS. Our next step is an SSD cache for ZFS, and then as prices continue to tumble, we'll switch to all SSDs everywhere.

        That 50k iops were measured in what? 4k operations? 16k? What? I could claim I can pull 1 million amps out of my house socket, which would be true... as long as the voltage is 0.0018v. IOPs are just as meaningless.

        SSDs are better. But so are Ferraris. Neither make much business sense outside of a drag race. Yea, if you have one small DB that's getting hit a LOT, then yes, they can make a difference. But in most operations upgrading to SSDs would costs millions and provide little benefit other than allow yo

        • I'm curious if data centers are moving to SSDs.
          Also, what about hosting companies, for high traffic web sites they host. I could see this as a premium service.
        • by fnj (64210)

          No, sorry, you couldn't get a million amps out of your wall socket if you turned the voltage down to 0.0018v. First, a million amps would incinerate your #14 copper wire and set fire to the house. Second, IR losses in #14 wire at such an absurdly low voltage would prevent the current from reaching anywhere near that high.

          Similarly, you obviously don't understand how IOPs relate directly, for example, to database performance, and how pretty much any SSD is 100 to 1000 times the speed of any rotating disk dri

        • Re:Ye Gods, an Ad (Score:5, Informative)

          by mcrbids (148650) on Monday June 02, 2014 @08:31PM (#47151567) Journal

          IOPs are anything but meaningless. For any kind of performance computing, they are one of the most commonly unrecognized bottleneck.

          IOPS is simple: how many random seeks can your storage device perform? If you can scootch your heads to the starting sector once per second, you have 1 IOP. Divide the rotational speed of your drive by 60. EG: 7200/60 = 120. That's the literal maximum number of seeks you can get out of your hard disk heads assuming that there is no seek time.

          The "k of operations" is irrelevant when discussing IOPS.

          How an idea so simple could be so commonly misunderstood is beyond me. It's true that IOPS won't matter if you are streaming a single, large media file. It's equally true that you can't serve more than about 120 random seeks in a second on a 7200 RPM drive. This is disguised a bit because your OS will try to minimize the seeks and aggregate seeks that are similar and/or close together.

          SSDs are now only about 5x the cost of HDDs in many cases. In past years, it's typical to have, multi-disk arrays solely to improve performance. In these cases, a single SSD can be not only dramatically faster, but significantly cheaper to boot.

          • by ttsai (135075)

            IOPS is simple: how many random seeks can your storage device perform? If you can scootch your heads to the starting sector once per second, you have 1 IOP. Divide the rotational speed of your drive by 60. EG: 7200/60 = 120. That's the literal maximum number of seeks you can get out of your hard disk heads assuming that there is no seek time.

            Hmm, if you assume there is no seek time, then ideally IOPS will be 120 if your request size is exactly a full track, which is almost never the case. IOPS will be 1s / (seek_time + rotational_latency + access_time + overhead). How many IOPS you can get from a 7200rpm drive depends on a lot of factors. Are the requests uniformly randomly distributed? Is NCQ enabled? What is the max track span of the requests? If the max track span is narrow and on the outer part of the disk, then you can get a lot IOPS

          • How an idea so simple could be so commonly misunderstood is beyond me. It's true that IOPS won't matter if you are streaming a single, large media file.

            You really need to consider fragmentation is a major issue on spindle drive performance.
            SSD's have no issue with fragmentation simply because their IOPS is so high and seek time is near 0ms~.
            IOPS does kind of matter :)

            Wonder how much of 2 minutes can be cut for loading up a VS2013 project, on one of these bad boy SSD's.

        • I know a number of people who make use of virtualization on notebooks, and SSDs help dramatically there. I switched to an SSD on my home system and since then, it's become painful being on any system with an HDD because of the latency caused by the drive. I'm trying to talk my boss into letting me get an SSD for my work notebook as I usually have at least one VM running and often two, and the competition for the hard drive is killing me.

          It's not a necessary thing for every person who has a notebook, but i

        • That 50k iops were measured in what? 4k operations? 16k? What? I could claim I can pull 1 million amps out of my house socket, which would be true... as long as the voltage is 0.0018v. IOPs are just as meaningless.

          In general, real-world usage, a good rule-of-thumb is 100:1 speed up for random seeks when comparing SSDs to 7200 RPM drives. Maybe only 50:1 for 15k SAS drives.

          Since enterprise SSDs are only about 2x-3x the cost of the equivalent sized 15k SAS drive, you have to ask whether that 50x-100x im
      • Perhaps it's an ad, but it's one that interests me.

        I'd be more interested if they paused now in their efforts to make them faster, higher capacity, or cheaper (they're in the very practical range now), and instead started concentrating a lot more on making them last longer.

        • I think maybe the idea is that in 5 years when it wears out the new tech available will be so much better you'd want to replace the drive anyway.

      • by arth1 (260657)

        We switched our DB servers to SSDs and saw over 90% reduction in average query latency.

        Yes, but was average latency the problem, or worst case write latency?

        The latter seldom goes down with SSDs, especially not for database applications with frequent random writes. Unless you place the SSDs on RAID controllers with a goodly amount of RAM, and don't care about atomic commits, SSDs can make the averages look oh so much better, but at the same time inject second long waits at the worst possible moments, while a drive has to erase an entire sector before a write can happen.

        I find SSDs to be abs

        • by arth1 (260657)

          Errata: s/\(mostly-\)write/\1read/1

          (But I guess everybody understood that from the context anyhow.)

        • by Anonymous Coward

          Are you sure about that? It is in random writes SSD has it's STRENGTH compared to HDD, simply because of the lack of seeking. I would have absolutely no issue with putting a write-heavy DB on SSD, but transaction logs would go to single spindle HDD, which gives better sequentual writes, and does not meet the wall when the SSD has used up all of it''s "erased" space. SSD's don't really write directly, first, they have to erase, then they write. Only when it is completely saturated, and there are no more free

      • If you are looking for an even bigger performance jump, upgrade to pci-e based ssd. Works very very well for databases and for certain types of workloads. Fusion-io, Intel, and dell sell enterprise versions, but many other vendors are rapidly getting into this segment as well.

      • SSDs don't have this problem. 50,000 IOPS is "no big deal" for an SSD, meaning that even if you have 40 million tiny 10k documents, you can still saturate your 6 Gbit SATA interface with sweet, sweet data.

        SATA. Bleurgh. How do you do multi-initiator with SATA?

        If it's not SAS I can't see any use for it.

    • by timeOday (582209)
      If this goes much further I could imagine seeing articles about the iPhone 6 on news sites someday!
    • by doccus (2020662)
      Could be.. since the "special pricing" seems to be on par with all the rest...
  • You should only trust SSD's produced by flash manufacturers:

    Intel, Samsung, Sandisk, and Micron.

    The first 3 use a company-wide branding strategy, whereas Micron keeps their product brands separate from their manufacturing brand. Crucial is Microns SSD brand.

    Micron seems to have got one up on the competition right now with regard to flash technology.
    • by Guspaz (556486)

      Micron doesn't have any advantage over Intel when it comes to flash technology, since they both get their flash from IMFT: Intel-Micron Flash Technologies.

      The advice to only get SSDs from flash manufacturers is sound, though.

      • by TWX (665546)
        You mean that SSD that I bought from the scruffy-looking fellow in the parking lot might not be legit?
    • by Krishnoid (984597)

      Toshiba [phys.org] also belongs to this club, but they only recently seem to be making SSDs available to the masses.

      • They also recently bought the SSD part of OCZ which had a reputation for being one of the shittiest SSD makers around. I'd want to hold off for a while to see if they can turn that around before buying their SSDs.

    • by vux984 (928602)

      Intel, Samsung, Sandisk, and Micron.

      I've had nothing but success from Mushkin SSDs; and they generally seem very well reviewed; Is there something wrong with them I should know about?

      • by Krishnoid (984597)

        Mushkin is a pretty reputable name, but if they ended up going with a less-reliable source, they could blame SSD failures on the flash manufacturer.

        A large well-known flash manufacturer trying to point the finger for SSD failures would damage the reputation of their own flash memory and SSD divisions, so my assumption is that they wouldn't release a consumer product at all if it would put them in that position.

  • by Oysterville (2944937) on Monday June 02, 2014 @06:15PM (#47150685)
    I have the Disable Ads box checked. Technical glitch?
  • Is that .43 cents or dollars per GiB?

  • wwooo (Score:2, Funny)

    by binarylarry (1338699)

    I'm going to go buy 100 of these right now because I read this on slashdot, who Okolona buddy!

  • by Guspaz (556486) on Monday June 02, 2014 @06:30PM (#47150791)

    The prices are good, but they're not much cheaper than existing drives; the Samsung 840 EVO 1TB goes for $450, or $0.45/GB.

    Micron's advantage is that they're using MLC, while the 840 EVO is using TLC.

  • Or do I need to buy a spinning disk to back up to?

    I've bought two Intel SSDs so far, an 80 and a 160. The 80 was lightly used (I hope, heh) and the 160 was an unused pull, it has a weird identity but it behaves fine. I figured I could trust them even with eBay provenance, and SFSG.

    Crucial, I'm less confident in. Is that justified? Or is this really just faith-based purchasing?

    • by binarylarry (1338699) on Monday June 02, 2014 @06:36PM (#47150833)

      You should never trust any drive and always backup your data.

      • by drinkypoo (153816)

        You should never trust any drive and always backup your data.

        Yeah, I back up all my data that I care about to a couple of different disks. But only if I truly don't trust a disk do I bother to have another one online right next to it set up as the next disk in the BIOS and mirror to it periodically. Slow laptop drive pulls are good for that sort of thing, and indeed it's what I'm using in my current Win7 games system, with a 160GB SSD mirrored to a 160GB 2.5" HDD.

      • by Aryden (1872756)
        +1 if I had mod points.
      • You should never trust any drive and always backup your data.

        Can you trust the backup drive?

    • by gman003 (1693318)

      I have a laptop with both a Crucial M4 and a regular hard drive. It's been going strong for two years now (well, the display is dying but that's not really relevant). Going off history, I expect the hard drive to die first, but I admit that's a completely unscientific prediction.

      You were right to be cautious, though - back in the early days of SSDs, there were many that were absolute crap (OCZ drives had horrible failure rates, and JMicron controllers were rubbish performance-wise). Intel was really the onl

    • by jon3k (691256)
      Dumbest question of the year - "Should I backup my hard drive?"
  • by epyT-R (613989)

    I wouldn't want to depend on this as my usage patterns tend to kill ssds prematurely. It's nice they're getting cheaper though.

  • Newegg routinely discounts [newegg.com] the Kingston V300 120GB SSD to $60 if you watch out for it (currently at $75 as of this posting). Why pay $80 when you can pay $60 for the same size and performance? If this post is an ad, it kinda sucks.
    • by PRMan (959735)
      I have a Kingston and other than the time that the room got up over 110 degrees F (and it lost all its data) it's been fine. Great, actually.
      • by harrkev (623093)

        Kingston 480G is $250 right now on the Egg. The 512G crucial is $225.

        Sorry, but the Crucial drives have good reviews from the hardware sites, and are actually quite cheap for the capacity. 1/2 terrabyte of solid state for well under $250? Yes, please.

    • Because $79.99 is the list price before Newegg discounting.

    • by citizenr (871508)

      Scaminston V300?

      http://www.anandtech.com/show/... [anandtech.com]

      Half the speed at same price? count me in!11

      • Thanks for that, it was interesting. It also explains why one of the V300 SSDs I bought in a batch for a set of desktop builds tested out at half the raw read speed when the standard hard drive diagnostic was performed. (Still pretty darn fast though.)
  • Have had terrible experiences with Marvell. I know, anecdotal. YMMV.

    • by SeaFox (739806)

      Have had terrible experiences with Marvell. I know, anecdotal. YMMV.

      That stands for "your Marvell may vary" in this case, right?

  • I still have a conventional disk in my desktop rig. I also frequently end up with a web page with Flash that nearly brings the entire comp to it's knees, due to near-constant disk activity. It's bad enough my electro-mechanical drive is being worn down by garbage software; not sure I could live with an SSD being literally consumed by the same indefensible cause.
  • but here's my observation .43 * 128 != 79.99 so, right off the bat, whatever this drivel is may be assumed a lie, garbage or both.

In every non-trivial program there is at least one bug.

Working...