• Posted in
• Storage
• SSDs
• Samsung
• TLC
• V-NAND
• 3D NAND
• 850 EVO

Endurance: Close to Planar MLC NAND

The big question with every new NAND generation is the endurance. We already saw6,000 P/E cycles in the SSD 850 Pro andan amazing 40,000 P/E cycles in the SSD 845DC Pro, which proved that V-NAND provides substantially better endurance over today's planar NAND nodes. However, endurance was never really an issue with planar MLC NAND except in the enterprise space, so the 850 EVO with its TLC V-NAND offers a much more interesting insight to the capability of 3D NAND technology.

To test endurance, I put the 120GB 850 EVO through our usual endurance test suite. Basically I just used Iometer to write 128KB sequential data at queue depth of 1 to the drive while monitoring the Wear Leveling Count (WLC) and Total LBAs Written SMART values. The 'Current Value' of the WLC SMART value gives the remaining endurance as a percentage (starts from 99), whereas the 'Raw Data' value indicates the number of consumed P/E cycles. In order to estimate the endurance, I had to find the spot where the increase in 'Raw Data' value decreases the 'Current Value' by one.

It appears that TLC flavor of V-NAND is rated at about 2,000 P/E cycles. The raw WLC value seems to be based on the user capacity (i.e. 120GB = 1 P/E cycle) because just going by it puts the endurance at ~2,133 P/E cycles (128/0.06), but that doesn't add up with the raw NAND capacity and total data written. However, the estimated total write endurance (which is just 15,260/0.06) suggests that the NAND itself is rated at 2,000 P/E cycles, which would make sense as the number of P/E cycles is usually an even thousand and it's also inline with the increase that the 850 Pro saw (from 3,000 cycles in the 840 Pro to 6,000 cycles).

While write endurance in client workloads was never truly an issue even with planar TLC NAND, the doubled endurance in TLC V-NAND makes it practically impossible to wear out the drive before it has become totally obsolete. Only some very extreme workloads could wear out the smaller capacities before the warranty runs out, but the 850 EVO is a wrong drive for such workloads in the first place. All in all, there should be absolutely no reason to worry about the endurance of the 850 EVO, especially given the endurance ratings Samsung is giving to the 850 EVO (75TB for 120/250GB and 150TB for 500GB/1TB).

97 Comments

View All Comments

• TheWrongChristian - Monday, December 8, 2014 -link

  Performance consistency wise, the MEX controller on the 1TB drive looks better than the MGX controller on the smaller drives. I guess this is the loss of a cpu core at work, but I figure there'll be no discernible difference to the user experience.
  
  All in all, looks like a nice drive, a reasonable upgrade to the existing 840 evo drives. Just hope the V-NAND cost brings the whole price down to compelling levels as the process matures.

• Solid State Brain - Monday, December 8, 2014 -link

  As for the Wear Leveling Count, it might actually be taking into account that a fixed portion of the installed NAND is used in SLC mode for the TurboWrite buffer. The 120GB model has 128GiB of NAND, of which 9 GiB are used for 3 GiB TurboWrite Buffer, so that makes 119 GiB of TLC capacity for both overprovisioning and user addressable space.
  
  By the way, this also implies that because of TurboWrite these Samsung EVO SSDs (including the previous 840 EVO) have less overprovisioning space than most other SSDs.

• Solid State Brain - Monday, December 8, 2014 -link

  Yikes! This was not meant to be in response to TheWrongChristian.

• hojnikb - Monday, December 8, 2014 -link

  >What I do know is that Samsung started the mass production of TLC V-NAND later, which suggests that the two aren't completely uniform. Moreover, from what I know TLC NAND requires some changes to the peripheral circuitry in order to read three bits from one cell, so while the NAND memory arrays could be alike the die size is still likely at least slightly different.
  
  Is it possible, that samsung designed 2nd gen 3D with TLC in mind (eg requred peripheral circuitry) and simply set the controller in 850PRO to use 4 states instead of 8 (so MLC). I mean, it kinda makes sense to go this way, but not the other way around....

• Solid State Brain - Monday, December 8, 2014 -link

  @Kristian Vättö
  Do you have any information from Samsung about whether the TurboWrite SLC buffer also helps decreasing the write amplification like on the SanDisk Ultra II with nCache 2.0?

• Kristian Vättö - Tuesday, December 9, 2014 -link

  Unfortunately I don't. It was among the questions I sent but unfortunately Samsung couldn't get any of my questions answered on time for the review.

• Solid State Brain - Wednesday, December 10, 2014 -link

  Too bad that you couldn't get an answer in time. According to my observations from other people's drives, it really looks like the Turbowrite does help on that regard. Try checking out my thread on the Memory and Storage forum.
  
  This phenomenon (write amplification just over or below 1.0x) is likely not going to show up during reviews or heavy usage since generally the drives get secure erased often and/or get hammered with writes which end up filling the Turbowrite SLC buffer.

• Daniel Egger - Monday, December 8, 2014 -link

  It would be nice to have some older drives in the charts to get some perspective on whether a drive would make a good upgrade choice. For an upgrade the available space and the space/price ratio are probably the most important aspects but some features or a major speed increase might sweeten the deal even more. ;)

• sheh - Monday, December 8, 2014 -link

  Not perfect of always complete, but usable:
  http://www.anandtech.com/bench/SSD/65

• Kvaern - Monday, December 8, 2014 -link

  It's my anecdotal experience that all these awesome benchmark figures means absolutely nothing for the average user.
  Case in point I upgraded from an old 60GB corsair drive to an EVO840 which on paper is like twice as fast as the Corsair but in reality my user experience is exactly the same.