Sign in with
Sign up | Sign in

What's Important: Steady State Performance

The OCZ RevoDrive 3 X2 Preview: Second-Gen SandForce Goes PCIe
By

Steady State Performance

SSD manufacturers prefer us to benchmark drives the way they behave fresh out of the box because solid-state drives slow down once you start using them. If you give a SSD enough time, though, it will enter into a steady state, where performance starts to reflect more consistent long-term use. In general, reads are a little faster; writes are slower, while erases are the slowest yet.

An Extreme Example of Steady StateAn Extreme Example of Steady State

We want to move away from benchmarking SSDs fresh out of the box because you only really get that performance for a limited time. After that, you end up with "steady state" performance until you perform a secure erase and start all over again. Now, I don't know about you, but we don't reinstall Windows 7 on our workstations every week. So, while performance right out of the box is interesting to look at, it's really not that important in the grand scheme of things. Steady state performance is what ultimately matters.

While this is a new move for us, IT professionals have long used this approach to evaluate SSDs. That's why the consortium of producers and consumers of storage products, Storage Networking Industry Association (SNIA), recommends benchmarking steady state performance. It's really the only way to examine the true performance of a SSD in a way that represents what you'll actually see over time.

There are multiple ways to get to a SSD’s steady state, but we're going to use a proprietary benchmark storage benchmark from Intel. This is a trace-based benchmark, which means that we're using an I/O recording to measure relative performance. Our trace, which we're dubbing Storage Bench v1.0, comes from a two-week recording of my own personal machine, and it captures the level of I/O that you would see during the first two weeks of setting up a computer.

Installation includes:

  • Games like Call of Duty: Modern Warfare 2, Crysis 2, and Civilization V
  • Microsoft Office 2010 Professional Plus
  • Firefox
  • VMware
  • Adobe Photoshop CS5
  • Various Canon and HP Printer Utilities
  • LCD Calibration Tools: ColorEyes, i1Match
  • General Utility Software: WinZip, Adobe Acrobat Reader, WinRAR, Skype
  • Development Tools: Andriod SDK, iOS SDK, and Bloodshed
  • Multimedia Software: iTunes, VLC


The I/O workload is somewhat moderate. I read the news, browse the Web for information, read several white papers, occasionally compile code, run gaming benchmarks, and calibrate monitors. On a daily basis, I edit photos, upload them to our corporate server, write articles in Word, and perform research across multiple Firefox windows. 

The following are stats on the two-week trace of my personal workstation:

Statistics
Storage Bench v1.0
Read Operations
7 408 938
Write Operations
3 061 162
Data Read
84.27 GB
Data Written142.19 GB
Max Queue Depth
452


According to the stats, I'm writing more data than I'm reading over the course of two weeks. However, this needs to put into context. Remember that the trace includes the I/O activity of setting up the computer. A lot of this information is considered one-touch, since it isn't accessed repeatedly. If we exclude the first few hours of my trace, the amount of data written drops by over 50%. So on a day-to-day basis, my usage pattern evens out to include a fairly balanced mix of read and writes (~8-10 GB/day). That seems pretty typical for the average desktop user, though this number is expected to favor reads among the folks consuming streaming media on a larger and more frequent basis.

On a separate note, we specifically avoided creating a really big trace by installing multiple things over the course of a few hours, because that really doesn't capture real-world use. As Intel points out, traces of this nature are largely contrived because they don't take into account idle garbage collection, which has a tangible effect on performance (more on that later).

Display all 37 comments.
This thread is closed for comments
  • 3 Hide
    reyshan , June 28, 2011 10:26 AM
    YoT!damn fast ssd and damn expensive ssd. might buy one 5 years from now(pci-e kind).
  • 2 Hide
    KingArcher , June 28, 2011 4:15 PM
    Wish I had won a lottery :) 
    So that I could afford me some drives like these.
  • 6 Hide
    Why_Me , June 28, 2011 4:29 PM
    omg this drive is fast! But way out of my budget :( 
  • 4 Hide
    warmon6 , June 28, 2011 4:30 PM
    Quote:
    It's not a business-class product. It's for the power user who is able to tax it using the right workload. If you're not one of those folks, the RevoDrive 3 X2 is seriously overkill.


    OVERKILL?!?!

    Nothing is overkill in the computer arena in terms of performance. :p 

    Just the price can be over kill. o.0
  • 7 Hide
    julius 85 , June 28, 2011 4:34 PM
    Just the price can be over kill. o.0

    For me the price is a bottleneck :) 
  • 7 Hide
    ElectroGoofy , June 28, 2011 4:39 PM
    Dear Santa...
  • 2 Hide
    acku , June 28, 2011 4:42 PM
    Santa is going to need a bigger expense account... :) 

    Personally, I'm hoping that OCZ adds TRIM prior to September.

    Cheers,
    Andrew Ku
    TomsHardware.com
  • -1 Hide
    chefboyeb , June 28, 2011 5:21 PM
    Jesus!
  • -2 Hide
    greenrider02 , June 28, 2011 5:43 PM
    I saw defense of the Vertex 3's occasional low numbers, but no mention of the solid (and sometimes better) performance that the cheaper and more miserly Crucial m4 showed throughout your tests.

    Perhaps you have some bias towards the Vertex 3 that needs reconsideration?

    Other than that, $700 seems like a fair price when considering the performace difference, especially if utilized properly, for instance as a high traffic web/corporate server
  • 0 Hide
    acku , June 28, 2011 5:51 PM
    greenrider02I saw defense of the Vertex 3's occasional low numbers, but no mention of the solid (and sometimes better) performance that the cheaper and more miserly Crucial m4 showed throughout your tests.Perhaps you have some bias towards the Vertex 3 that needs reconsideration?Other than that, $700 seems like a fair price when considering the performace difference, especially if utilized properly, for instance as a high traffic web/corporate server


    If you read the first page then you know that I give a nod to Vertex 3s as the fastest MLC based 2.5" SSD. I consider that plenty of love. :) .

    We'll discuss the lower capacity m4s in another article. FYI, I suggest that you read page 5 and page 6. We are not testing FOB. We are testing steady state. That's part of the reason the SF-based drives are behaving differently with incompressible data.

    On your second point, this is in no way targeted toward an enterprise environment (that's what Z-drives are for). There is no redundancy in the array if a single SF controller fails. The whole card is a dud afterward. You can add higher level redundancy, but enterprise customers have so far been nervous on SandForce products. Plus, there's a general preference for hardware vs. software redundancy. (That's them talking not me). Overall, this makes it unacceptable for any enterprise class workload.




  • 0 Hide
    Supertrek32 , June 28, 2011 5:54 PM
    You know, almost any application that would actually benefit from these speeds won't be very effective on these drives. Why? Not enough storage space.

    It's like having a car that can do 300 miles an hour, but can only carry enough fuel to go 20 miles. Does it have a niche? Yeah. Is it practical? Not really.
  • 0 Hide
    cronos177 , June 28, 2011 6:13 PM
    In a couple of years I see them for like 1/2 the price. ONLY worth it for people who earn their paychecks based on the ability to finish task faster. that's the sole purpose at the moment.
  • 1 Hide
    guzami77 , June 28, 2011 6:16 PM
    I have the X2 100GB... im not that impressed. The speed isnt meeting specs, and other hardware/software doesnt always like a PCIe(non-standard) hard drive. Also dont forget you cant overclock your PCIe voltage with one of these...
  • 0 Hide
    acku , June 28, 2011 6:18 PM
    guzami77I have the X2 100GB... im not that impressed. The speed isnt meeting specs, and other hardware/software doesnt always like a PCIe(non-standard) hard drive. Also dont forget you cant overclock your PCIe voltage with one of these...


    I really recommend that you update the firmware. I do notice a speed difference. If you've seen our compressible benchmarks, then you know that over time the RevoDrive X2 can perform poorly. The lack of TRIM doesn't help. :( 
  • 1 Hide
    chefboyeb , June 28, 2011 6:20 PM
    The prices are too steep tho...
  • 0 Hide
    compton , June 28, 2011 6:36 PM
    This is an excellent exploration of SSD performance in general. Without the benefit of storage benches I myself have been wondering why one of my SSDs seems so much faster that the other, when it seems at face value that the answer should be clear.

    As far as OCZ goes, people who can benefit from this drive are a small group, but I'm not exactly sure who that is. Sure, its fantastically fast, and faster is better, but most people would tragically under-utilize a product like this. Since it's clearly not for enterprise use, I just have to guess what a typical user looks like for this device. Is it the high end media creation freelancer? Someone who makes their bones with Photoshop? Who knows? I want one, even though I'd clearly be better served by a more pedestrian drive. It's clearly destined for the "Cool Wall".
  • 2 Hide
    acku , June 28, 2011 6:43 PM
    Ok. I feel like I need to clear up some confusion. Some sites are reporting TRIM support and we're saying there is no TRIM support.

    The RevoDrive 3 X2 has hardware support for TRIM through VCA 2.0. This much is true, but you can't use TRIM because of a software problem. As for why?

    OCZ has a problem getting the TRIM command to the RevoDrive 3 X2, because it uses SCSI commands over PCIe, hence StorPort SCSI drivers. The TRIM command is out of the question because it's technically part of SATA. The only alternative is Unmap, which is to SCSI what TRIM is to SATA. Unfortunately, Windows does not support Unmap as part of its native driver stack. Furthermore, TRIM commands are only issued by Windows 7 when you empty the Recycle Bin, but you must have your SSD set to AHCI (part of SATA). Obviously this different from SCSI. So effectively no TRIM. :( 

    We have been in contact with the Windows driver team and OCZ, so we know that a fix is being explored. If it materializes, that's another matter....

    (RevoDrive and RevoDrive X2 don't don't support TRIM at the hardware level, so the issue is moot for them.)

    Cheers,
    Andrew Ku
    TomsHardware.com
  • 2 Hide
    Niva , June 28, 2011 6:47 PM
    Ideal user would run massive databases that need tons of accesses all the time. Are there drivers for linux for this? Small business server with the right usage will make this worth it. Also the 'trim' issue should be a non-issue under linux whereas win 7 apparently doesn't include the unmap command.
  • 2 Hide
    christop , June 28, 2011 9:05 PM
    Crazy fast!!!
  • 0 Hide
    warmon6 , June 28, 2011 9:45 PM
    __-_-_-__$700?! wow that's a nice price. Though performance @1.25gbps is kind of low for a pci-e solution. there's already on the market +2gbps. anyway those cost +$7000!I'm going to wait and see if there's any defect with this ssd's...


    Where you get Gbps (= Gigabits) from? ;)  http://en.wikipedia.org/wiki/Data_rate_units#Suffix:_b_vs_B

    Were talking about GBps (Gigabytes). :p 
Display more comments