Like all NAND Flashes, these in turn are based on memory cell technology, which has undergone a great evolution as an architecture since their first electronic circuits and although we will not explain in depth how they work. or what is their story, yes. will give a brief review to focus on the current events of the CLQ as such.
Why Choose NAND Flash Over Common RAM Chips
One of the paradoxes of SSD vs system RAM is that although the former are slower, less dense and more expensive to manufacture than the latter, it is true that the technologies and the approach they are designed for completely separate them, although its principles are the same in reality.
The main difference is that NAND Flash as a technology is based on the FGMOS or MOSFET with floating gate, which are a type of metallic semiconductor field effect transistor. This implies that the Gate as such is isolated and with direct current, which is the basis of the technology of an SSD: The objective is that each cell maintains its electrical seal, thus making it possible to save the data which is stored.
RAM doesn’t work like that, but being MNV or non-volatile memory type, it always needs a current to flow through it, so its architecture is based on MOS capacitors. Therefore, the SSD seeks to store each bit of information so that the controller, CPU, or GPU can access it and read each cell based on 0 and 1.
Therefore, and of course, the first NAND Flash cells started out by being able to store 1 bit each in two different states (zero and one). This type of NAND flash memory was called SLC or Single Level Cell and is the basis of any old or current SSD, although with many nuances along the way.
A trend change in the industry with QLC
The principle of switching technology is simple, but it has both problems and advantages. What the storage industry wants is to increase the capacity of these devices as a primary goal, but behind that is speed, price per GB included and of course durability, for example.
Here is the whole controversy, so we will explain everything about QLC memories so that it is understood. As we have already guessed, QLCs are fourth level memories or Four-level cell and they assume that they are based on the technology introduced with its predecessors, the FTAs.
These types of third-level memories achieve performance similar to current MLCs, but with shorter lifecycles, typically and on average between 3000 and 5000 cycles per cell. This means reducing its useful life, but due to its new structure, they manage to lower the price per GB, so the durability is acceptable for over 10 years seen in a consumer SSD, it is a technology. valid and very balanced in all respects. . . .
From SLC to MLC, via TLC and now QLC, what are the technological innovations?
With QLC, the industry manages to increase the capacity, but also the complexity and durability of each cell. You get 4 bits per cell, but up to 16 different voltages must be implemented, one for each state for the number of bits contained. Therefore, this amount of states and voltages results in a loss of reading speed and especially writing to an SSD.
The validation and verification of each bit is also more complex and each cell degrades much more, so neither the ETC Common is sufficient to maintain structural integrity and other error checking methods had to be used. Logically, if one has to resort to all this, the internal read and write cycles are higher (not to be confused with the external data transfer cycles themselves) which has an impact on a duration of total life between 500 to 1500 erase, write and read cycles.
Therefore, the use of SLC caches has become necessary to increase the performance of these SSDs to levels of at least TLC. Ergo, as the industry takes the industry up a notch with more bits per cell, speed and durability drops at the expense of capacity and that’s where the whole problem lies.
The market is trying to switch to QLC, but at the moment neither the costs are lower when comparing the TLC vs QLC SSD models on their own, nor the pros outweigh the cons at the moment.
It is true that there are models where the optimal circumstances occur to claim that QLC is the same or better, but these are niche models where users trust and this generates a higher volume of chips and therefore a cost. inferior.
So why choose QLC over a good SSD with TLC? Mainly because of capacity and costs. Whenever SSD comes out at a price lower than GB with increasingly higher yields and with greater capacity, so it is only a matter of time that TLC remains marginal in the market. Likewise, with QLC a new type of NAND Flash called X-NAND que paradójicamente puede arrebatarle el trono de la industria a la primera, puesto que abarca parte de los problems inherentes a esta tecnología y que a su misma vez mantiene los estándares de los niveles de celdas, por lo que podría comenzar con QLC y seguir con PLC no problem.
The future of QLC is bright, slow yes, but the industry is increasingly widening the gap between new levels, allowing technology to mature and manufacturers to compete. API it will be its successor and while there is great progress with it, it is really a long way off in the market to see high performance consumer SSDs in generous capacities at competitive prices.