Any idea how much the loss in endurance will be compared to 19nm MLC flash? It probably will not affect us in daily use (only for very heavy write loads), but it's an interesting question that is worth looking at.
Another observation, the 15nm 128Gb chip is not small enough to fit in a microSD (15x11x1mm with some room for the plastic shell.sp i am assuming very max dies size would be 9x14 so 126mm2) and since the current 128GB is 16 layers plus the controller in a 1mm thick package, it seems unlikely to see bigger microSD from Sandisk ,or anyone else, soon.
Well, they could go with bigger than 64Gb but smaller than 128Gb MLC or some TLC that is small enough but i'm pretty sure nobody is considering 4 bit on such an advanced process , maybe they could on 3D if they use an old process but we don't have details on what process they'll use for 3D and i got the feeling it won't be that old.
In September we'll see the 128Gbit 3-bit version of Samsung's 32-Layer, 3-D NAND. This is the same die as the 95mm2 86Gbit 32-Layer die which stores 2-bit per cell. The density increases to 1.35 Gbit/mm2. http://www.businesskorea.co.kr/article/5785/3bit-v...
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
12 Comments
Back to Article
CrazyElf - Friday, August 22, 2014 - link
Any idea how much the loss in endurance will be compared to 19nm MLC flash? It probably will not affect us in daily use (only for very heavy write loads), but it's an interesting question that is worth looking at.Kristian Vättö - Friday, August 22, 2014 - link
It should be rated at the same 3,000 P/E cycles -- at least that has been Toshiba's/SanDisk's goal with every die shrink.Infy2 - Friday, August 22, 2014 - link
I wonder how the durability and cost of manufacture compare to Samsung's new 3D NAND.hojnikb - Friday, August 22, 2014 - link
Wow, thats significantly better than Microns solution. Hopefully smaller die size will yield in lower street prices for SSDs.Are they having any plans to move TLC to 15nm aswell ?
Kristian Vättö - Friday, August 22, 2014 - link
15nm TLC production is also ramping up, though it could take a bit longer for it to reach SSD maturity (usually ~2-3 quarters).hojnikb - Friday, August 22, 2014 - link
Great!So i guess its realistical to expect 15nm TLC ssd year or so from now ..
jjj - Saturday, August 23, 2014 - link
Interesting that Micron's 128Gb die shrinks only by some 14% from 20nm to 16nm.Any clue what size is the second gen 19nm 128Gb die?
jjj - Saturday, August 23, 2014 - link
Another observation, the 15nm 128Gb chip is not small enough to fit in a microSD (15x11x1mm with some room for the plastic shell.sp i am assuming very max dies size would be 9x14 so 126mm2) and since the current 128GB is 16 layers plus the controller in a 1mm thick package, it seems unlikely to see bigger microSD from Sandisk ,or anyone else, soon.jjj - Saturday, August 23, 2014 - link
Unless they use smaller TLC dies and that would be likelyhojnikb - Saturday, August 23, 2014 - link
Or even 4 bits per cell, which was also used by sandisk some time ago.jjj - Saturday, August 23, 2014 - link
Well, they could go with bigger than 64Gb but smaller than 128Gb MLC or some TLC that is small enough but i'm pretty sure nobody is considering 4 bit on such an advanced process , maybe they could on 3D if they use an old process but we don't have details on what process they'll use for 3D and i got the feeling it won't be that old.Hans de Vries - Saturday, August 23, 2014 - link
In September we'll see the 128Gbit 3-bit version of Samsung's 32-Layer, 3-D NAND. This is the same die as the 95mm2 86Gbit 32-Layer die which stores 2-bit per cell. The density increases to 1.35 Gbit/mm2. http://www.businesskorea.co.kr/article/5785/3bit-v...