Sign in with
Sign up | Sign in

MTTF? MTBF? "My Drive Lasts Longer"

OCZ's Vertex 3: Second-Generation SandForce For The Masses
By

I wanted to cover reliability because I've spent some time on a few forums (including those dedicated to storage), noticing that people are discussing MTBF ratings. This deserves a separate page of discussion.

When you look at the data sheet of any drive, you'll notice reliability is expressed in mean time between failures (MTBF) or mean time to failure (MTTF). These two values share the same units. The only difference is that the first assumes you can fix a drive, and in latter, replace it.


SF-1222 SF-2141 SF-2181 SF-2281 SF-2282 SF-1565 SF-2582 SF-2682
Target MarketClientClientClientClientClientEnterpriseEnterpriseEnterprise
MTTF (hours)
2 million
2 million2 million2 million2 million10 million10 million10 million


If you look at the mean time to failure (MTTF) of any enterprise controller from SandForce, such as the SF-2582, you'll notice that it's rated at 10 000 000 hours. Yet, the SF-2281 is rated at 2 000 000 hours.

If you do the math, you realize that 10 000 000 hours roughly equals 1140 years. Does this mean you can bequeath an enterprise drive to your tenth-generation progeny?

Not at all. This rating is based on probability. There are many ways to calculate MTBF, but here is one example. Let's say SandForce had 2000 drives based on the SF-2582 in its qualification lab. If you were to turn on all of these SSDs at the same time and start the clock, every passing hour would equal 2000 hours of drives running. For the sake of this example, let's assume the first SSD fails after 2000 hours (~82 days). You would stop the clock the moment the first drive failed and the MTBF would be calculated based on the number of drives originally set up. Because only one drive out of 2000 failed after 2000 hours of use, the MTBF would be four million hours.

The problem with this type of math is that it is hardly considers advanced statistics or stochastic math. For all we know, the entire batch of 2000 drives could have failed five minutes after we stopped the clock. This is why MTBF can be an inflated number (depends on the math used).

While the math may seem a bit odd, this relates to the MTBF rating you see on a drive. For example, Corsair’s Force SDDs use the same SF-1222 seen on OCZ's Vertex 2. Yet, Force drives are rated at 1.5 million to the 2 million hours for Vertex 2. This occurs because the MTBF rating of a drive is based on a sum probability. There are many components that go into making a drive: memory, buck converters, controllers, resistors, and so on. Even though the SF-1222 is rated for 2 million hours, the parts that Corsair is using collectively add up to 1.5 million. Since any system is only as strong as its weakest link, this results in a lower MTBF rating. OCZ claims a higher MTBF because the rating it gives to the parts other than the SandForce controller exceed 2 million hours. Hence, OCZ publishes the reliability rate of the controller as that of the drive.

We point this out because MTBF is not a meaningful way for us to measure how long a drive will last. At the end of the day, it would be incorrect to compare the durability of two different drives simply using the MTBF rating, even if they use the same controller (like the Force and Vertex 2). If you love math, you've already come to the realization that this number can be easily manipulated. The fastest way is to simply increase the number of drives being tested. Double the drives, double the MTBF. Let's put aside the flaw in this approach to calculating MTBF for a moment. Without knowing how many drivers were originally tested, you can't determine how long a single drive might last. Remember this when you are shopping for drives, be it disk or solid state.  (If you like academia as much as I do, I recommend reading a paper written by Bianca Schroeder and Garth Gibson for more information. Eduardo Pinheiro, Wolf-Dietrich Weber and Luiz Andr´ Barroso at Google Research have also written a great paper that explores reliability of consumer drives.)

Ask a Category Expert

Create a new thread in the Reviews comments forum about this subject

Example: Notebook, Android, SSD hard drive

Display all 57 comments.
This thread is closed for comments
  • 0 Hide
    lradunovic77 , February 24, 2011 12:41 PM
    SSD prices are really down slow and that is because HDD is still selling.
  • -4 Hide
    lradunovic77 , February 24, 2011 1:17 PM
    I agree and it is time for HDD to be retired. We don't need them anymore, but Servers.
  • 3 Hide
    bildo123 , February 24, 2011 1:18 PM
    A far cry as far as "the masses" are concerned. Still, over $2/GB is too much. Getting closer however. I'd pay $200 for a 256GB SSD with these speeds.
  • 0 Hide
    lradunovic77 , February 24, 2011 1:19 PM
    I am sure price will really drop by end of this year.
  • 0 Hide
    acku , February 24, 2011 1:20 PM
    mayankleoboy1the fact that they use ~15% of a quad core SB CPU, is amazing.with the mechanical drives, they were just sitting idle. this more than anything, makes the SSD worthwhile.


    Well what I didn't mention in the review is that the benchmark starts as ~20% across all cores during the first 10 seconds, which is from PCMark setting up the disk trace. After that, the IO activity throttles a single core up to 100% for almost all SSDs. For the hard drives, we see ~60-80% utilization of a single core.

    Cheers,
    Andrew Ku
    TomsHardware.com
  • -2 Hide
    lradunovic77 , February 24, 2011 1:23 PM
    I say keep your desktop active all the time. I am running i980x overclocked to 4.0Ghz and there is no way i will put my computer into any type of power saving mode, it is useless and power saving is just mimick. We are talking about very small amount of money over a year. Having Turbo option makes sense from certain point of view but bottom line is that it is just wasted silicon and pretty much useless.
  • 0 Hide
    vvhocare5 , February 24, 2011 1:24 PM
    "The problem is that any price above $2/GB is going to be a hard sell unless you're an early adopter by nature. Our choices in recent System Builder Marathon stories reflect this. Look at our December $1000 PC."

    Overall a good article. Anyone into MTBF's will find that one page uninforming and anyone not into it is likely lost.

    I would disagree with that statement only in the sense that a $1000 PC is not going to be filled with high-end superior performing parts. So I dont see a reason to apologize for its price. The person who can afford a $3000+ PC isnt going to blink buying the 240G model and will likely see it as entirely reasonable.

    Me? I think I have found my next ex-drive.....

  • 0 Hide
    acku , February 24, 2011 1:35 PM
    Quote:
    so as i said, will OC increase the scores a bit?

    and what about power saving enabled?


    None of our tests were executed in an environment that allowed any idling. Furthermore, we disabled CPU throttling. Power saving was enabled in the sense that the display was allowed to turn off, which is part of the default profile in Windows.

    OCing may increase performance, but only to the extent that the bandwidth will support it. As I mentioned, PCMark throttles a single core up to 100%. It isn't a sustained trend.
  • 0 Hide
    bto , February 24, 2011 1:39 PM
    on your 1000 dollar gaming system, I'd rather have a vertex 3 than two 460's hell even an agility 2... and still afford better than a 460.
  • 0 Hide
    bto , February 24, 2011 1:42 PM
    But then that's me, and to quote the great Inigo Montoya, "I hate wait" and most games I play are not bleeding edge, I also work on my computer, play HD movies and copying time makes me angry when I'm moving files. Buy an ssd, and later spend 60 bux on a 1tb drive down the road.
  • 0 Hide
    Anonymous , February 24, 2011 1:45 PM
    There is a minor typo in the vertex 3/ vertex 3 pro comparison table - sequential read performance appears twice instead of sequential read and sequential write (the accompanying text sorts it out though in the end).
  • 0 Hide
    acku , February 24, 2011 1:49 PM
    Quote:
    There is a minor typo in the vertex 3/ vertex 3 pro comparison table - sequential read performance appears twice instead of sequential read and sequential write (the accompanying text sorts it out though in the end).


    Sorry about that. Fixed!
  • 1 Hide
    acku , February 24, 2011 1:56 PM
    Quote:
    on your 1000 dollar gaming system, I'd rather have a vertex 3 than two 460's hell even an agility 2... and still afford better than a 460.


    I completely understand that sentiment. Keep in mind that 460 prices dropped after our December System Builder Marathon. The fact remains that a decent CPU along with a SLI or CrossFire config will likely put the kibosh on any SSD choice. With a gaming rig, you are going to want a performance SSD, which means those lower end options almost go out the door. In any case, if you end up choosing a graphics card that runs at least $200, you'll hard to fit a SSD into the budget. It isn't impossible the way prices are falling, but you will likely be relying heavily on rebates or special deals.
  • 0 Hide
    acku , February 24, 2011 2:00 PM
    Quote:
    @ Andrew Ku

    thanks for the clarification. :) 


    No problem! We're here to answer any question.
  • 0 Hide
    bto , February 24, 2011 2:16 PM
    ackuI completely understand that sentiment. Keep in mind that 460 prices dropped after our December System Builder Marathon. The fact remains that a decent CPU along with a SLI or CrossFire config will likely put the kibosh on any SSD choice. With a gaming rig, you are going to want a performance SSD, which means those lower end options almost go out the door. In any case, if you end up choosing a graphics card that runs at least $200, you'll hard to fit a SSD into the budget. It isn't impossible the way prices are falling, but you will likely be relying heavily on rebates or special deals.


    Forgot to mention: Thank you for the article!
    I love SSD info I used to have 4 of the gigabyte I-RAM cards that were bootable in RAID 0 to get some similar results (two I-RAM were double stick taped to my case wall with pci flexible riser cards to make them all fit. I was hooked long before that obviously. Silence is also a golden feature that you do not realize until you have an already quiet system and remove the HDD.

    also to quote myself: "I hate Rebate" as probably everyone else does.
    Even the Vertex 2 60GB is down to $105 so for $35 more (which now you'd probably save more than $35 getting same components) You get a good ssd, later (if immediate budget did not permit) you can drop in a 1-3 TB hdd for storage. Since you already have 60GB it can wait a paycheck or two. replacing main drive later with SSD requires imaging which doesn't always work well, well at least it didn't used to. I had to try three different imaging programs before it came across properly (acronis, ghost, active disk image) I have an idea also for a fun article. Build an extremely small pc. Possibly with a 1.8" SSD to go in a car. with mii itx it would be doable and with ssd you'd not have to worry about the HDD crashing over speed bumps. Or do an article of a mini itx that lan party gamers could stick on the back of their monitor. I've replaced a case door on a silverstone with a 24" samsung before (wiring it in and all) like an Imac.
  • 1 Hide
    binoyski , February 24, 2011 2:52 PM
    If there was a time to save my money for a pc build, it was the time that Sandy Bridge launched. OMG! LGA 2011 will be the "GOLDEN AGE" for pc enthusiast!

    - A superfast SSD
    - Thunderbolt(Lightpeak)
    - 4-8 cores Ivy Bridge CPUs

    And I just hope that Bulldozer will come out w/ a BANG so prices will be lower!
  • -1 Hide
    notsogosu , February 24, 2011 3:00 PM
    Am i the only one this article and then rundown to anandtech to read their version? :S
  • 2 Hide
    restatement3dofted , February 24, 2011 3:07 PM
    notsogosuAm i the only one this article and then rundown to anandtech to read their version? :S


    Probably not, but so far, you're the only person that seems to be impressed enough with themselves to post about it...
  • 0 Hide
    LordConrad , February 24, 2011 3:59 PM
    I love how they boast about AES 256-bit protection. I'm no expert, but considering the recent attacks on AES 256, I think think AES 128 is more secure.

    http://eprint.iacr.org/2009/317
  • 0 Hide
    alidan , February 24, 2011 5:01 PM
    thanks for incorporating the SATA 3Gb/s configurations. i think there were some interesting numbers, at least when considering that while in 6Gb/s the drives preformed better, but when moved to 3Gb/s other 3Gb/s drives preformed better, still a bit tired from just waking up, ill go through it again later to see if i missed a reason for this happening.
Display more comments