This huge capacity SSD with every bit handled by few combinations of transistors with limited life will be really making the life of the chip unpredictable. How much will be the real life of these products. Remember the Floppies then CDs and DVDs loosing data. In this case when the company is claiming the 5 years life, what will be the actual life of the a bit stored.
This title is completely misleading. Micron will be at 16 nm in the 4th quarter, Samsung will be at 19 nm eventually, and SK Hynix has not ever given a time period for when they will be at 16 nm. The article is about Skyera, which is not at all addressed in the title.
It appears prior to Samsung's V-NAND announcement, everyone was planning for 16 nm. Now, the question is whether 16 nm will be dropped for the 3D rollout. Micron already came out with 16 nm, I think it means their 3D NAND is further off. Since 16 nm requires more than double patterning, I am still surprised Micron is saying it is lower cost per bit. It goes directly against the common impression of the prohibitive cost of any multiple patterning.
A possibility is that Micron, maybe SK Hynix and others, moved to quadruple patterning in the low 20's, perhaps one layer at a time. So by the time they reach 16 nm, the cost difference from 19 nm is minimized.
I agree. Optical media was once touted as the more reliable solution to data storage but I have found that many 10 year old RW optical disks have failed and are essentially absolete. The only reliability approach seems to be to continue to by larger hard drives as data accumulates and simply replace older drives with newer ones. I suppose that this would work for SSD's but this process tends to be price sensitive and hard drives have managed to stay ahead in the $/GB realm..
Blog Doing Math in FPGAs Tom Burke 2 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...