This year, the first solid state drives based on 3D QLC NAND memory chips appeared. This memory provides storage of four bits of data in one cell, which means higher data storage density, and consequently lower cost of solid state drives. However, in reality everything turned out to be a bit more complicated. https://uaeu.top
Less than a month ago, Intel introduced the first consumer solid state drives for 3D QLC NAND memory, called Intel SSD 660p. The new products offer good performance, and the official cost of the 1 TB model is only $ 200. And one would expect that the appearance of QLC storage devices from other manufacturers will reduce the cost to an even lower mark.
However, according to industry sources, this will not happen, at least in the current generation of memory chips. The fact is that at the IMFT joint venture between Intel and Micron, the percentage of 64-layer 3D QLC NAND chips suitable for use in solid state drives is only 48%. And that’s very little, considering the fact that the percentage of usable 64-layer TLC chips (three bits per cell) at IMFT is about 90% today.
It is not surprising that when launching new technologies, manufacturers face a high percentage of marriages. Usually companies work on this, and over time the percentage of marriage decreases. However, in the case of IMFT, the manufacturer is more busy completing the development of promising 96-layer flash memory, and therefore resources to debug 64-layer QLC may not be enough.
Thus, the economic effect of a 33 percent increase in storage density, which in theory should provide memory with QLC-cells, is completely neutralized by production difficulties. Moreover, given the yield of suitable NAND crystals, the transition to QLC is not yet able to give any gain in the cost of solid state drives. And this may mean that the emergence of affordable high-volume SSDs, which was expected with the spread of QLC 3D NAND, may not happen immediately, and the role of QLC-memory in the process of falling prices for solid state drives will be minimal.