Sign in with
Sign up | Sign in

PCMark 8's Storage Consistency Test: New For Tom's Hardware

Plextor M6S And M6M SSD Review: Revving Another Marvell Engine
By

Futuremark's PCMark 8 takes everything that was good about PCMark 7's storage testing and makes it better. The standard PCMark 8 storage components include real trace activity for 10 different activities. These recorded I/O segments include productivity, gaming, and photo/video manipulation, yielding individual sub-tests. At the end, each category is scored in seconds, a master throughput score is conferred (similar to the average data rate in our Tom's Hardware Storage Bench), and the benchmark generates a total score.

In preparation for a transition away from PCMark Vantage and PCMark 7, I started testing PCMark 8 with Plextor's M6e. You'll see it appear in an upcoming review. But for now, this is what that drive looks like under PCMark 8 normally:

PCMark 8 Storage

PCMarks
Bandwidth
WoW
Battlefield 3Photoshop (Light)Photoshop (Heavy)Adobe Illustrator
M6e M.2 PCIe 256 GB
4971
280.57 MB/s
58.4s
133.4s
113.7s
359.9s
70.8s
PCMark 8 Storage

Adobe InDesignAdobe After EffectsAdobe IllustratorMicrosoft WordMicrosoft ExcelMicrosoft PowerPoint
M6e M.2 PCIe 256 GB57.5s70.8s
72.0s
28.3s
9.2
9.2

Although this is illuminating, the newest version of PCMark 8 Professional includes extended storage tests that let us dig deeper into SSD performance.

In truth, on a new or lightly-used drive, it's hard to distinguish one model from another. It takes tough workloads over long periods of time to expose the issues that might affect our recommendations. PCMark 8's storage consistency component is designed to hit SSDs with those workloads, getting consumer-oriented products dirty enough to sink the bad ones. 

Whether you run PCMark 8's normal or extended tests, those 10 traces form the backbone of this benchmark. As they're played against the drive under test, information is measured and recorded. The extended storage tests simply repeat the traces multiple times using varying conditioning procedures before each session.

First, the raw block device (there is no partition) is preconditioned twice by filling the entire accessible LBA space with 128 KB sequential writes. Once that is completed, the first degradation phase randomly writes blocks between 4 KB and 1 MB in size to random LBA spaces on the drive. Since the writes aren't 4 KB-aligned much of the time, the SSD's performance drops quickly. After all, non-4 KB-aligned accesses create overhead and generally increase write amplification significantly.

The first Degradation Phase begins with 10 minutes of those punishing random offset writes, after which each PCMark 8 activity trace is played against the SSD being tested. The successive degradation rounds are similar, except an additional five minutes are tacked onto each iteration. After eight repetitions, that write period expands to 45 minutes.

Next comes the Steady Phase. Each of five Steady Phases writes 45 minutes worth of random offset data prior to trace playback, pushing the drive even harder and making it more difficult to perform housekeeping duties. With fewer blocks available for writing, latency increases substantially.

Lastly, PCMark 8 moves into a Recovery Phase, which consists of five idle minutes before trace playback. Repeat that five times, and the test concludes.

This particular PCMark 8 test is brilliant in that it stresses SSDs methodically with preconditioning, followed by trace playback. You end up with a ton of data covering bandwidth, latency, and duration for all 10 traces and each of the 18 phases. But tracking that information lets us paint a picture of drive performance through each step of the benchmark. Every drive takes a beating during the Degradation and Steady Phases, but the Recovery Phase should push the most resilient drives to the forefront. SSDs that rely on garbage collection during write activity (and not background garbage collection) may not benefit much from this recovery time. 

Sounds like fun, right? With a bit of background out of the way, let's take our first look at the results.

Ask a Category Expert

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

Example: Notebook, Android, SSD hard drive

Display all 10 comments.
This thread is closed for comments
  • 0 Hide
    bjaminnyc , April 10, 2014 6:47 AM
    I know there are slight performance differences between models and manufacturers, for the home user the discrepancies matter very little other than lifespan. It all comes down to price point for non-commercial application, not sure where this really fits in the product stack.
  • 0 Hide
    WithoutWeakness , April 10, 2014 8:21 AM
    The Marvell 88SS9188 controller in the M6S is only capable of addressing 4 NAND channels at a time whereas the 88SS9187 (Crucial M500, Plextor M5Pro) and 88SS9189 (Crucial M550) are capable of addressing a full 8 channels. I would expect the Plextor M6Pro to use the newer 88SS9189 found in the M550 to be able to address 8 channels of NAND and support up to 1TB drives using 128Gbit NAND.
  • 0 Hide
    Ipatinga , April 10, 2014 8:42 AM
    Clean looks... ok update for the M5S... here hopping M6P will have something on its sleeve performance wise. I liked the Plextor software (its like a bonus for some people... ).

    But I still have the Samsung 840 EVO on my list, after all, when the dust settles, they both perform very well in a notebook for a regular joe, but the EVO has the price to beat.

    Since my Crucial m4 is not dead... I will wait... mainly for prices to go even lower so I can get a higher capacity model instead of the best of the best.
  • 0 Hide
    JOSHSKORN , April 10, 2014 12:41 PM
    Why not compare the Samsung EVO 840?
  • 0 Hide
    cryan , April 10, 2014 1:51 PM
    Quote:
    The Marvell 88SS9188 controller in the M6S is only capable of addressing 4 NAND channels at a time whereas the 88SS9187 (Crucial M500, Plextor M5Pro) and 88SS9189 (Crucial M550) are capable of addressing a full 8 channels. I would expect the Plextor M6Pro to use the newer 88SS9189 found in the M550 to be able to address 8 channels of NAND and support up to 1TB drives using 128Gbit NAND.



    So we weren't able to get confirmation from Plextor or Marvell, but it's probably the case that the 9188 is a cut down four channel version of the 9189. It's all the more annoying, because this is the third review in a row with new Marvell silicon but no actual info from Marvell.

    We have a pretty good idea of this from looking at the 9174 4 channel used in the UltraPlus vs the 8 channel 9175 used in.... pretty much everything else ever made... that the 9188 is a four channel version of the 9189. But we don't know how many CEs per channel it has, or much else for that matter.

    However, we do know Plextor's M6 Pro will have a 1 TB version, also with A19 flash. Plextor reps told me as much at the 2013 Flash Memory Summit. That was last August though, so I think there have been some delays -- either with Marvell, Toshiba, or both.


    Regards,
    Christopher Ryan
  • 0 Hide
    cryan , April 10, 2014 1:58 PM
    Quote:
    Why not compare the Samsung EVO 840?


    If you're referring to the 128 KB sequential and 4 KB random performance, I'm trying to keep those clean. I'd much rather throw down on matchups like this in more important metrics. However, with TRIM testing out this go around, the EVO v. M6 angle got downplayed. It's an oversight.

    Regards,
    Christopher Ryan
  • 0 Hide
    Zeroplanetz , April 10, 2014 6:37 PM
    I really wish write speeds would actually make a jump.
  • 0 Hide
    cryan , April 10, 2014 9:51 PM
    Quote:
    I really wish write speeds would actually make a jump.


    For sequential writes, you can't do much better than 500 MB/s. The fastest M6M/M6S hits 440 MB/s, and so there just isn't much room left to grow unless you switch to PCIe-based storage or add a couple drives in RAID.

    Regards,
    Christopher Ryan

  • 0 Hide
    michaelrhaley , April 16, 2014 6:38 AM
    Great article! I'd like to see a summary comparison of your top ten drives listed in Tom's best SSDs for the money using your new benchmark. Most important is your omission of Samsung 256G Pro which has been your leader for 8 months cited now as "face meltingly fast". Any performance benchmark must include the Samsung leader. Love to see this right away from you!
    Sincerely, Michael
  • 0 Hide
    kalmquist , May 4, 2014 10:15 PM
    Currently, the 128GB M6S is selling for $100 on Newegg, the 256GB M6S is selling for $165. The 512GB M6S is not listed.

    Hard to justify buying these over the Samsung 840 EVO ($85 and $154). Actually, for most uses my first thought would be the Crucial M500 ($75 and $120). The M500 doesn't win on either performance or power consumption, but it is good enough that these will not be major issues for the typical user.