Samsung TLC NAND write endurance study part 1

I just posted the article Samsung TLC NAND write endurance study part 1.

In part one of this study, I’ll be looking at the write endurance of TLC NAND.
TLC NAND is not new. It has been used in USB flash drives for a while, where write endurance is not really an issue.
Samsung launched the first consumer SSD with TLC NAND not long ago, in the shape if the 840 series.
Does TLC NAND have what it takes for SSD use?
Let’s make a start to finding out in part 1 of this article.

Click to read the full article here: [](

Feel free to add your comments below. 

Please note that the reactions from the complete site will be synched below.

I’ve followed that thread for a long time, Wischmop, These tests are not even remotely based on any real life usage pattern, nor take many factors into consideration.
if I wanted to kill any SSD in a number of days, I could have easily run the same tests, and proved nothing. :slight_smile:

As a data point, my Intel SSD 330 series 180 GB drive (MLC) in my laptop has a little over 8 TB writes over a period of 10 months, or about 27 GB writes per day.

I don’t use the laptop at the office but apart from that it’s my main machine and the SSD has been used a lot for encoding movies (DVD/TV/BD) into MP4/MKV format, which counts for a lot of those writes.

Media Wear Indicator of my drive is still at 100%.

^ Likewise for my SandForce based SSDs (MWI 100%).
They should be of course, MLC has at least twice the endurance when compared to TLC, and with SF controllers being able to compress data on the fly, they have probably the lowest write amplification of any SSD controller.

Follow the leader.

Hi HighEndToys,

Could you explain your comment please? We welcome everyone to share their views or criticize us, we can only learn from that…

PM coming.

To prove something, a single SSD can´t do that for a large number of drives :wink:

Maybe u have a good, maybe you have a bad drive, maybe something in the middle :wink:

Still think dead flash-memory isn´t the main reason for defect drives (except maybe OCZ with old Indilinx-Controller, very ugly wear-leveling and real cheap NANDs). Many defect SSDs aren´t recognized anymore, the read-only-mode is very rare.So I think it´s the controller, maybe bad firmware.

The other point is, if it survives one year, maybe this type of SSD is EOL. Then the next drive have other controller, other NAND-finishing (15/16nm)

I have two 840 units here. Never worried about longevity. One internal 120GB, one external USB 3.0 250GB.

Well, it depends on how much use write on it.

Are you worry about your other SSDs :wink: ?