TurboWrite: MLC Performance on a TLC Drive

All NAND trends towards lower performance as we move down to smaller process geometries. Clever architectural tricks are what keep overall SSD performance increasing each generation, but if you look at Crucial's M500 you'll see that it's not always possible to do. Historically, whenever a level of the memory hierarchy got too slow, the industry would more or less agree to insert another level above it to help hide latency. The problem is exascerbated once you start talking about TLC NAND. Samsung's mitigation to the problem is to dedicate a small portion of each TLC NAND die as an SLC write buffer. The feature is called TurboWrite. Initial writes hit the TurboWrite buffer at very low latency and are quickly written back to the rest of the TLC NAND array.

Since the amount of spare area available on the EVO varies depending on capacity, TurboWrite buffer size varies with capacity. The smallest size is around 3GB while the largest is 12GB on the 1TB EVO:

Samsung SSD 840 EVO TurboWrite Buffer Size vs. Capacity
  120GB 250GB 500GB 750GB 1TB
TurboWrite Buffer Size 3GB 3GB 6GB 9GB 12GB

I spent some time poking at the TurboWrite buffer and it pretty much works the way you'd expect it to. Initial writes hit the buffer first, and as long as they don't exceed the size of the buffer the performance you get is quite good. If your writes stop before exceeding the buffer size, the buffer will write itself out to the TLC NAND array. You need a little bit of idle time for this copy to happen, but it tends to go pretty quickly as it's just a sequential move of data internally (we're talking about a matter of 15 - 30 seconds). Even before the TurboWrite buffer is completely emptied, you can stream new writes into the buffer. It all works surprisingly well. For most light use cases I can see TurboWrite being a great way to deliver more of an MLC experience but on a TLC drive.

TurboWrite's impact is best felt on the lower capacity drives that don't have as many NAND die to stripe requests across (thus further hiding long program latencies). The chart below shows sequential write performance vs. time for all of the EVO capacities. The sharp drop in performance on each curve is when the TurboWrite buffer is exceeded and sequential writes start streaming to the TLC NAND array instead:

On the 120GB drive the delta between TurboWrite and standard performance is huge. On the larger drives the drop isn't as big and the TurboWrite buffer is also larger, the combination of the two is why the impact isn't felt as muchon those drives. It's this TurboWrite buffer that gives the EVO its improvement in max sequential write speed over last year's vanilla SSD 840.

Endurance: Not a Problem Even at 19nm RAPID: PCIe-like Performance from a SATA SSD
POST A COMMENT

133 Comments

View All Comments

  • Touche - Thursday, July 25, 2013 - link

    I don't think your tests are representative of most people's usage, especially for these drives. TurboWrite should prove to be a much better asset for most, so the drive's performance is actually quite better than this review indicates. Reply
  • Sivar - Thursday, July 25, 2013 - link

    Really well-written article.
    I have to admit, while most of Samsung's products are crap, their 840 and later SSDs are not bad at all.
    (The 830, while not prone to electronic failure, was built really poorly. It's SATA connector would snap off if you tilted your head the wrong way while looking at it).
    Reply
  • Coup27 - Thursday, July 25, 2013 - link

    Samsung have gotten into the world position they are in today by selling crap. I have used plenty of 830's and I have never had an issue with the SATA connector so I have no idea what you are doing with it. Reply
  • Coup27 - Thursday, July 25, 2013 - link

    Haven't ^^ (why is there no edit button?) Reply
  • piroroadkill - Thursday, July 25, 2013 - link

    So you accidentally broke a SATA connector, and now that's suddenly a flaw? I have two Samsung 830 256GB in my system, and somehow I didn't break the SATA connectors...
    I also fitted 4x Samsung 830 256GB to a server at work.. and somehow I didn't break the SATA connectors..
    Reply
  • HisDivineOrder - Saturday, July 27, 2013 - link

    True, this. SATA connectors are poorly designed, but that's the fault of the people who made the spec, not the specific one in the 830. I'm not saying it can't break. I've had SATA connectors break on a variety of devices. None of them were my 830, but I'm not saying it's impossible or whatever.

    I've seen WD, Seagate, and Hitachi drives all have a problem with the connector, though. Seems like SATA and HDMI were designed to make the connection as loose and easily broken as possible. I guess that gives them some small percentage of people buying all new product to replace something on said product that's small and plastic...
    Reply
  • mmaenpaa - Thursday, July 25, 2013 - link

    Good article once again Anand,

    and very good perfomance for this price range.

    Regarding Torx, I believe this is one the main reasons why it is used:

    "By design, Torx head screws resist cam-out better than Phillips head or slot head screws. Where Phillips heads were designed to cause the driver to cam out, to prevent overtightening, Torx heads were designed to prevent cam-out. The reason for this was the development of better torque-limiting automatic screwdrivers for use in factories. Rather than rely on the tool slipping out of the screw head when a torque level is reached, thereby risking damage to the driver tip, screw head and/or workpiece, the newer driver design achieves a desired torque consistently. The manufacturer claims this can increase tool bit life by ten times or more"

    (from https://en.wikipedia.org/wiki/Torx)

    BR,
    Markku
    Reply
  • hybrid2d4x4 - Thursday, July 25, 2013 - link

    For what it's worth, my experience with screws is consistent with your post. I've never had a torx screw slip out, which is definitely not the case with philips or the square or flathead varieties. I'd like to see them used more often. Reply
  • piroroadkill - Thursday, July 25, 2013 - link

    Agreed. I love Torx. Philips and pozidriv are the terrible bastard children of the screw universe. Always slipping and burring. Ugh. If everything was replaced with totally cam-out free designs like Torx, allen head, robertson screw.. etc, etc.. then I'd be more than happy. Reply
  • psuedonymous - Thursday, July 25, 2013 - link

    I'd LOVE for Torx to be used more often. They're much easier to work with (not once have I had a Torx screw fall off the screwdriver and roll under the desk), the screwheads are more robust, and they frankly look a lot nicer than Philips or Pozidriv.

    It'd make pulling apart laptops all day a darn sight less onerous if Torx were the standard rather than Philips.
    Reply

Log in

Don't have an account? Sign up now