TRIM and Performance over Time

When I first tested the C300 I noted that it fell victim to the same bug that plagued Intel’s X25-M: throw a bunch of random writes at the drive and it’ll get backed into a corner that it’ll never get out of. The scenario that you have to create to force the drive into this condition isn’t realistic at all, however that isn’t to say that over months or years of heavy use that you wouldn’t find yourself in the same situation. Either way, Crucial corrected the problem and spent weeks validating the firmware fix. Crucial did continue to sell drives during this period, which I wasn’t exactly happy about.

Either way the problem is now fixed with the 0002 firmware revision. It’s a destructive firmware update so you’ll lose any and all data on your drive after applying it. I also had a strange issue where the update had to be applied twice on my 256GB drive before it would work properly. Here’s what the C300’s performance looks like after being hammered for 20 minutes with random writes:

Ouch. Writes drop down to a dismal 20MB/s. But now if we TRIM the contents of the drive the performance curve looks like this:

Writes are at about 140MB/s and reads are in the low 200s. Everything is back to normal.

Now note that unlike the SandForce drives, Crucial’s controller/firmware combination will let the drive get into this state. Even if I throw randomly generated data at a SandForce drive the performance remains very high (the drive in the chart below is OCZ's Vertex 2):

Note that the lower line actually represents read performance - write speed is still up above 180MB/s.

The C300’s performance is simply not as resilient as what SandForce has to offer. For the majority of users this isn’t a problem. Typical desktop workloads shouldn’t get the drive into this state. Add in the efficiency of Crucial’s TRIM implementation and you should never see such a drop. However if your workload is atypical or particularly write intensive, the C300 may not be a good fit.

The C300 does perform background garbage collection. I’ve never particularly been a fan of idle GC simply because it fixes a problem after it’s created rather than avoiding the problem altogether. If you don’t have TRIM however (running RAID, Mac OS X, any Windows OS other than 7), idle GC is all you can rely on.

How well does it work? I subjected my 128GB C300 to almost the same torture test I mentioned above. I filled the drive to 80% of its total capacity and then threw random writes all over the drive for 20 minutes straight. Performance was predictably bad:

I left 20% of the drive empty because free space is needed for the idle GC to take effect. Here’s what the drive’s performance looked like after being idle for 3 hours:

Not exactly a drastic improvement. The idle GC just isn’t very aggressive. Compared to what Indilinx and SandForce have done, Crucial’s implementation just doesn’t cut it. For Windows 7 users running a single drive this shouldn't matter since you have TRIM. But if you're running RAID or another OS, the C300 isn't as desirable.

Power Consumption Final Words
POST A COMMENT

51 Comments

View All Comments

  • Breit - Wednesday, July 14, 2010 - link

    yes, a sandforce drive should be slower in general when used with third party encryption as all data seems random to the drive then and the compression algorithms have nothing to compress. i'm not sure if the sf1200 controller can, but the sf1500 controller definitely can encrypt the data on the drive by themself storing it in an aes-128 encrypted format.

    quote from the sandforce site:
    "DuraClass technology automatically stores data in a secure, AES-128 encrypted format. This also prevents would-be thieves from extracting data directly from the flash memory should they ever have access to the drive."
    (http://www.sandforce.com/index.php?id=21&paren...

    question is how good or usable their implementation is?! :)
    Reply
  • sparkuss - Tuesday, July 13, 2010 - link

    I asked this question in a previous review and once again you note your "using the Marvell 88SE9128 controller similar to many motherboards".

    Did you only use the WIN7 MSAHCI and INTEL IMSM 8.9 on both the onboard SATA2 and Rocket 620?

    This is of interest to many on the Crucial boards concerning the lack of TRIM support in Marvell drivers for their 9123/9128 controllers.
    Reply
  • mattgmann - Tuesday, July 13, 2010 - link

    last year the ssd market was pretty cut and clear with what was good and what wasn't. There have been a lot of advances recently, and it's getting hard to keep track of what manufacturers with what controllers work well in what configurations. I'd really love to see an SSD buying guide with all of the current drives, possibly separated into workstation, laptop, and server use sections.

    thanks
    Reply
  • willscary - Tuesday, July 13, 2010 - link

    I just sent my first SSD back to Crucial. I called. They RMA'ed me and will send a replacement ASAP. I am impressed. I have 5 year warranties on the M225 series SSDs, and they never even blinked...send it back and they will replace it.

    This was the only M225 that I have installed that does not have 1916 firmware.

    My comment is on the prices. If any of you recall, I was the guy who already had bought, installed and used 5 Crucial M225 128GB SSDs when I decided to try Sandforce in the OWC SSD. After I bought the SSD and ON THE DAY IT SHIPPED, OWC changed their website and said that the controllers were Sandforce 1200 and not the 1500 that I was told when I puurchased. It was still a good deal, but I was upset with the bait and switch, so I refused delivery and had it returned.

    I then purchased a Crucial M225 for myself, this one the 256GB flavor.

    I want to say that I have had no problems. Peak reads reach about 240MB/s and average reads are about 215MB/sec with large files. Writes are less, about 175-180MB/sec for larger files. With small files like Anand uses, my writes are about 125MB/sec and my small file reads are about 145MB/sec.

    Very small files read and write much slower, in the 5-25MB/sec range, but those are files that are very small...in the 4-20KB range.

    What amazes me is the speed. While 5MB/sec may sound slow, it is not, at least when you consider that these are 4KB files. These small files write and read nearly instantly.

    I paid $525 for the drive. I have seen them as low as $509. This is a great price! Now that I have seen the newest article here by Anand, I really understand that my outdated M225 is still a great deal at $525.

    I would like to see this drive, or a comparable one, listed in the lineup with these newer drives. While not as fast, I believe them to be a good value when sale priced. I even saw a Western Digital 128GB SSD a few weeks ago for $219 with free shipping. They were gone very quickly!
    Reply
  • Makaveli - Tuesday, July 13, 2010 - link

    That was a good read thanks. Reply
  • pesos - Tuesday, July 13, 2010 - link

    Hi Anand,

    Thanks for the followup review - I have been using a 256gb C300 on each of my Poweredge T710 servers for the last few months with great results.

    I am curious - does applying the destructive firmware update restore the drive to peak performance, or should it be coupled with some kind of secure erase?

    Thanks,
    Wes
    Reply
  • elimliau - Wednesday, July 14, 2010 - link

    Hi,
    Has anyone made the above work, mine cant?
    Regards,
    elimliau
    Reply
  • sparkuss - Wednesday, July 14, 2010 - link

    elimliau,

    You need to head over to the Crucial SSD forum. I have seen several threads and posts on that specific card/board.

    http://forum.crucial.com/t5/Solid-State-Drives-SSD...
    Reply
  • sparkuss - Wednesday, July 14, 2010 - link

    EDIT: You may have to register to get that link to work, sorry Reply
  • foo-bar - Friday, July 16, 2010 - link

    This is all very interesting, and new SSDs seems to hit the market constantly. But the differences seems to be mostly software related i.e firmware. What are the bottleneck to get full SATA speed? Is it the NAND flash them selves? And if so, what is on the horizon when it comes to this type of memory? Reply

Log in

Don't have an account? Sign up now