Sign in with
Sign up | Sign in

New AS SSD Benchmark

Three PCI Express-Based SSDs: When SATA 6 Gb/s Is Too Slow
By

AS SSD, by Alexej Schepeljanski, is a benchmarking tool for solid state drives. It includes four synthetic tests plus final scores for read/write and combined results, and three real world copy tests.

The synthetic tests include sequential read and write operation, a 4 KB random read and write test, and the same thing using 64 threads. There is an access time test for read and write operation as well. The read access time test is executed over the full storage area of the device, while write access time test and the sequential and 4 KB random tests are limited to a 1 GB test file size. The benchmark returns three scores: one for read performance, one for write performance, and one combining both results. This is the weighting scheme:

  • Overall score = seq_write * 0.15 + seq_read * 0.1 + 4k_reads * 2 + 4k_writes + 4_64thrd_writes + 4_64thrd_read * 1.5
  • Read score = seq_read * 0.1 + 4k_read + 4_64thrd_read
  • Write score = seq_write *0.1 + 4k_write + 4_64thrd_write


You've probably noticed that the weighting for 4 KB random reads is much higher than the other areas. This is because the benchmark author, as well as different industry sources, consider this the most important performance metric.

The practical tests consist of a copy test suite that transfers an ISO image (two large files), a software package (one folder with many small files), and a game package (one folder with small and large files). The benchmark returns the effective throughput in MB/s as well as the time required to complete this workload. We like that this test uses Windows’s copy command, which effectively shows the drives’ capability of concurrent read and write operations.

We will be using AS SSD on future SSD reviews, but of course it will complement the existing benchmarks and not replace anything with the exception of h2benchw, which is not very suitable for testing SSDs that use caching.

The benchmark also supports a compression test run, which we will be using in future SSD reviews. It allows us to analyze how a drive copes with data that is well- or poorly-compressible. SandForce-based controllers rely heavily on compression, and there are interesting results to be found when you hit them with different types of data.

Ask a Category Expert

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

Example: Notebook, Android, SSD hard drive

Display all 32 comments.
This thread is closed for comments
  • 0 Hide
    amk09 , July 1, 2011 4:13 AM
    wow just wow, I want one so bad.
  • 1 Hide
    iubyont , July 1, 2011 4:55 AM
    I want to see this comparison updated when Intel's 720 SSD arrives. Were Intel's listed speeds mere exaggeration?
  • 6 Hide
    gkay09 , July 1, 2011 5:36 AM
    The OCZ Revo and the Revo X2 drives have been available for a long time also the X3 has just been released,... IMO these should have been reviewed here too,...
  • 2 Hide
    yukijin , July 1, 2011 6:30 AM
    i'm wondering how fast you can cold boot windows 7...
  • 1 Hide
    Jax69 , July 1, 2011 9:15 AM
    __-_-_-__you really should have tested iodrive OCTAL. that would kick ass this shitty pci-e ssd's. Also you should have compared with ramdisks.

    Ramdisks are orders of magnitude faster than any SSD available now. IoDrive is still very good despite it's age, the new ones are way faster.
  • 0 Hide
    shin0bi272 , July 1, 2011 9:54 AM
    that's all well and good but who cares?

    http://www.engadget.com/2011/06/30/embargo-ibm-develops-instantaneous-memory-100x-faster-than-fl/
  • 9 Hide
    Anonymous , July 1, 2011 10:38 AM
    First page: "To make a long story short, please don't take this review as a shootout, but as a look at different concepts and options."

    Last page: "Its performance is so impressive and so consistent across all of the benchmark categories that we have no choice but to declare the ioDrive this shootout's performance winner."

    What?
  • 1 Hide
    burnley14 , July 1, 2011 11:04 AM
    I'm actually not too impressed. For how much these things currently cost, they should be 10x faster than current SATA SSDs. It looks like more like 3x though. Still nothing to laugh at, but I think they have a long way to go still.
  • 0 Hide
    srgess , July 1, 2011 1:05 PM
    __-_-_-__you really should have tested iodrive OCTAL. that would kick ass this shitty pci-e ssd's. Also you should have compared with ramdisks.


    Ramdisk speed are like 7000 mb/s + if i remmember... I doubt ssd or pci ssd will come close to a loyal benchmark.
  • -2 Hide
    lradunovic77 , July 1, 2011 1:31 PM
    SATA has no future, eventually will be replace with PCIe in some other maybe form of connector.
  • -1 Hide
    Hupiscratch , July 1, 2011 3:14 PM
    Is not possible to install games in a Fusion-io board? I´ve always thought it was possible to install the games on the Fusion-io and the OS on a standard SSD.
  • 1 Hide
    biotkcdr , July 1, 2011 3:24 PM
    This test is seriously flawed. The LSI had 6 SSDs in RAID on one card. The ioDimms were stand-alone. They (ioDimms) scale linearly across all performance metrics as you raid them together. Stick an ioDuo drive in there and you will literally get twice the performance on a single slot (and beat your six raided drives in all metrics). Stick 2 duos in the PCIe bus and you will get 4x the performance. You can't do that with spinning disks and you can't do that with SSDs--you get diminishing returns and non-linearities. Can't boot? It doesn't matter so much. Keep all your caches, game files, applications, scratch disks, and indexes on the fusion cards and you'll be blown away. You should have pointed out how weak the LSI card was in reality. The larger capacity for that card was, in reality, multiple drives raided together on a single card. Big deal, you could do that with an ioOctal and have 8x performance over a single ioDimm. These other drives just don't stack up against the ioDimms. Agreed, the price point isn't for the enthusiast, but in an enterprise scenario, what you can do with an Fusion-io setup is simply amazing. Furthermore, you didn't even get into a discussion about how important latency is, or the wear life of the drives. Those are other places where Fusion-io is way way out in front FTW.
  • 1 Hide
    biotkcdr , July 1, 2011 3:31 PM
    shin0bi272that's all well and good but who cares?http://www.engadget.com/2011/06/30 [...] r-than-fl/


    Nice, but there are limitations on the theoretical limits of copper and current limitations on the pcie bus. It's all about how fast you can feed the processor and with how little latency. Yes, this is great, but it's how you deploy the technology on existing infrastructure that will matter (until that infrastructure catches up).
  • 1 Hide
    vvhocare5 , July 1, 2011 3:43 PM
    burnley14I'm actually not too impressed. For how much these things currently cost, they should be 10x faster than current SATA SSDs. It looks like more like 3x though. Still nothing to laugh at, but I think they have a long way to go still.


    I agree. The speed is interesting but not what it should be. Im thinking these cards also need a large RAM cache to make use of the PCI bus speeds and then let the SSDs work in the background -just like the high performance hard drive cards do. The controllers dont seem to be optimized for PCI-bus level speeds.
  • 1 Hide
    biotkcdr , July 1, 2011 3:58 PM
    No RAM cache is involved. You're either on the PCIe bus or not. Tom's has a diagram on one of these pages explaining the setup.
  • 0 Hide
    Apple Troll Master , July 1, 2011 4:50 PM
    lradunovic77SATA has no future, eventually will be replace with PCIe in some other maybe form of connector.


    LightPeak will replace SATA.
  • 0 Hide
    iwod , July 2, 2011 4:08 AM
    biotkcdrNice, but there are limitations on the theoretical limits of copper and current limitations on the pcie bus. It's all about how fast you can feed the processor and with how little latency. Yes, this is great, but it's how you deploy the technology on existing infrastructure that will matter (until that infrastructure catches up).


    Exactly. Not to mention in an non compressible truly random scenario, those Sandforce Controller wont be that much faster in Random Write. And all the funny things happen with Sandforce controller. Consistency is much more important. Fusion - IO truly wins this test with a technology that is now nearly 4 years old.
  • -2 Hide
    Anonymous , July 2, 2011 7:13 AM
    My PC boots up in 8 seconds to the password prompt. It's soooooo slowwwww..
    $110.00 for a 60GB SATA2 Mushkin. 240MB/sec both ways. Really: I love it!
  • 1 Hide
    sceen311 , July 2, 2011 2:19 PM
    "Three PCI Express-Based SSDs: When SATA 6 Gb/s Is Too Slow"

    Then why are you comparing these PCI Express SSD's to SATA @ 3 Gb/s?
    Seriously big flaw in your tests here.
  • -1 Hide
    fall0ut3 , July 2, 2011 9:52 PM
    This is crazy to think SATA 6 will be slow on SSD, even the cheapest ssd is a lot faster than a HDD
Display more comments