8

I recently got a Crucial M4 SSD for my mid-2009 MacBook Pro 13" (2.53 GHz). Since I paid some money, I would like to get the best performance possible out of it. Running a speed test using app store Black Magic Speed Test, it showed 260 Mb read and 170 Mb write, although on the Crucial website it says read up to 500 Mb. Trim is not enabled and firmware is the latest (0309).

The connection speed is SATA 3 GB/s.

Is this performance normal or something is going wrong?

From system profiler :

Vendor: NVidia
 Product:   MCP79 AHCI
 Link Speed:    3 Gigabit
 Negotiated Link Speed: 3 Gigabit
 Description:   AHCI Version 1.20 Supported

   M4-CT128M4SSD2:

      Capacity: 128.04 GB (128,035,676,160 bytes)
      Model:    M4-CT128M4SSD2                          
      Revision: 309
      Serial Number:    000000001204032BCB4A
      Native Command Queuing:   Yes
      Queue Depth:  32
      Removable Media:  No
      Detachable Drive: No
      BSD Name: disk0
      Medium Type:  Solid State
      TRIM Support: No
      Partition Map Type:   GPT (GUID Partition Table)
      S.M.A.R.T. status:    Verified
      Volumes:
    disk0s1:
      Capacity: 209.7 MB (209,715,200 bytes)
      BSD Name: disk0s1
      Content:  EFI
    Macintosh HD:
      Capacity: 127.18 GB (127,175,917,568 bytes)
      Available:    63.01 GB (63,009,751,040 bytes)
      Writable: Yes
      File System:  Journaled HFS+
      BSD Name: disk0s2
      Mount Point:  /
      Content:  Apple_HFS
    Recovery HD:
      Capacity: 650 MB (650,002,432 bytes)
      BSD Name: disk0s3
      Content:  Apple_Boot
6
  • 1
    +1 great question; I also have an M4 and want to know if I can make it even snapper Mar 13, 2012 at 1:00
  • posted on few places. wherever I get reply Ill post on the rest.
    – latusaki
    Mar 13, 2012 at 1:18
  • A factor I forgot is that 500 mb/s is for SATA 6gb/s. Would that double my read therefor getting around 500mb/s from current 260?
    – latusaki
    Mar 13, 2012 at 1:49
  • a 3gb/s connection should be able to give at least 0.375MB/s
    – Alexander
    Mar 13, 2012 at 2:01
  • 1
    Your stats are acceptable. I have a Vertex2 and it pulls the same numbers. While my Mac supports 6 GB/s throughput, the drive does not. Your problem is the other way around. To push upwards of 350+ mb, you'll need a Mac with a 6 GB/s controller. That is your bottleneck. The drive is functioning optimally under your current conditions.
    – user10355
    Mar 13, 2012 at 3:17

3 Answers 3

10

Your SSD's link speed is limited by the troughput of the SATA interface and the protocol overhead.

  • SATA 3 (6 Gbit/s): max. 600MB/s
  • SATA 2 (3 Gbit/s): max. 300MB/s
  • SATA 1 (1.5 Gbit/s): max. 150MB/s

A document by the Serial ATA International Organisation says:

What’s the real-world data transfer rate of SATA 6Gb/s?

Answer:
The realizable transfer rate across a 6Gb/s SATA link depends on the efficiency of the controller design on both the host and device sides of the interconnect. The SATA 6Gb/s interface transmits information at 600MB/s, however not all 600MB/s are realized as the user data payload because the protocol includes other data and handshaking communications between the host and device. In general, the SATA interface is very efficient. Realized transfer rates are typically very close to the theoretical maximum, which is one of the primary benefits of SATA technology for mass storage devices.

What overhead brings the 6Gb/s transfer rate down to the real-world throughput?

Answer:
There are two general categories of overhead that come into play: a communication used to send commands and receive status, and a low-level communication that handles handshakes between the host and the devices to assure the integrity of the transmission.

It's best to compare your throughput rates with that of other Crucial M4 users. An optimal comparison would compare disks using the same firmware, benchmark tool and SATA interface.

However, the best comparison I found is a thread on forum.crucial.com. In this thread users are comparing Crucial M4 128GB (firmware 009) speeds on a SATA 2 interface. Even though you are using a different firmware, the results should be somewhat comparable because the firmware update 0309 was not aimed improving transfer rates. Judging by this thread, I'd say that your throughput is fine.

The only way to get the promised speed of the Crucial M4 is by using a SATA 3 interface. In my MBP I'm currently using a Crucial M4 128GB (firmware 0309) with a SATA 3 interface. The Black Magic Speed Test gives me:

  • max. 510 MB/s read
  • max. 190 MB/s write
7
  • Thanks for the informative post. After lots of reading I decided to give trim enabler a try (on its latest release). Have you enabled trim ?
    – latusaki
    Mar 13, 2012 at 13:36
  • @latusaki I have never enabled TRIM during the last few months that I've been using the Crucial M4. During that time I did not notice a major performance loss. This German article indirectly mentions that the Firmware update 0009 has significantly improved the Garbage Collection in their test. I've even used the SSD with Filevault 2 enabled most of the time. But, I've decided to give TRIM a try like three hours ago, because it seems to be working fine for the Crucial M4. :)
    – gentmatt
    Mar 13, 2012 at 13:53
  • Ok.. if I get any indication that its a bad idea ill post.. hopefully crucial customer service will know for sure.
    – latusaki
    Mar 13, 2012 at 14:41
  • Reply from Crucial ended with this : TRIM however is a great tool and does help clean your drive for future use. If you can enable it then I would recommend doing so but if you can't then don't worry about it as Garbage Collection is doing the same job automatically.
    – latusaki
    Mar 14, 2012 at 15:58
  • 1
    @latusaki Here's a good read on TRIM and OS X. Your mileage may vary as your SSD doesn't have a SandForce controller onboard: apple.stackexchange.com/questions/25224/…
    – user10355
    Mar 21, 2012 at 9:40
3

You take a large hit running on an SATA2 rather than SATA3 connection. Taking a quick look at Anandtech's review of the M4, your numbers don't look out of line.

http://www.anandtech.com/show/4253/the-crucial-m4-micron-c400-ssd-review/3

The only easier way to make things snappier is to upgrade to a larger drive.

You need to be aware of performance over time. Since TRIM is not running, you are depending on the drive's garbage collection to maintain performance. There are ways to implement TRIM on Lion for non-Apple SSD drives but reviews are mixed. No idea how effective GC is on the M4 when using it with OS X.

4
  • 1
    Just a point to add to Bill's answer: Anandtech's article shows the drive's inboard garbage collection doesn't perform well at all. It is unadvisable to run TRIM on OS X should the drive have a SandForce controller, but the M4 doesn't. And would likely benefit from TRIM support. That's not an endorsement of TRIM on on OS X but a point of consideration.
    – user10355
    Mar 13, 2012 at 3:13
  • @cksum You're right, but I have to mention that this article is very old. This test was published on 3/31/2011 where the Crucial M4 had firmware 0001. There have been three firmware updates since then. 0002 on 06/8/2011, 0009 on 08/25/2011 and 0309 on 01/13/2012.
    – gentmatt
    Mar 13, 2012 at 7:06
  • The firmware update 0009 has "improved write latency for better performance under heavy write workloads". But I don't know what the performance increase has been like. I've not seen any tests using a crucial m4 with the firmware revision 0009 or 0309.
    – gentmatt
    Mar 13, 2012 at 7:08
  • @gentmatt Perhaps running Xbench after some time is the way to go. I know TRIM in OS X has no conflicts with anything but SandForce, but it may not be required in this case. Perhaps looking at performance after 6 months of using the drive and then taking action is the safest bet.
    – user10355
    Mar 13, 2012 at 9:16
0

The existing answers are good, but don't really mention one last thing which I will include here just to have it mentioned for completeness. Any speed ratings given for SSDs are an indication of the relative data transfer speeds of the controller built into the drive, not the actual drive. Now unlike a normal spinning platter drive where the controller stuff is largely on the motherboard and the drives are (caching aside) simply storing and retreiving your data, the controller on an SSD drive has more capabilities, and that is the ability to compress/decompress your data on the fly on the way in/out.

So, if you send over a carefully formatted data file that is easily (de)compressible then the controller will do so, and this may mean that ignoring the amount of read/write that is has actually done, it's figures may seem inflated. To get 500Mb/s you need these conditions. It's a bit confusing, but in order to present such a speed it won't actually have written at that speed, merely compressed and written (say it get's a 50% reduction/increase in size, then that's an effective doubling of data transfer).

Obviously this is a bit of chicanery, but they all do it. A more interesting stat are the IOP figures, but thats another story.

Black Magic is a truer test, because it uses data that cannot be compressed on the fly to increase transfer speeds as it already is heavily compressed. It gives a truer representation of your capabilities, and getting 250Mb + on a SATA2 line is actually pretty neat.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .