Don't be silly. Popular science: What's the difference between memory and storage?(2)
2. Unit of DRAM and NAND
As can be seen from the above storage principles, storage units of both DRAM and NAND are actually b. Why the capacity of storage product generally is marked with B, when the particle capacity of storage product is labeled with b.
Storage structure of DRAM memory particles is depth plus width. Let us take an official memory particles document of Micron as an example. The depth and width of memory particles numbered by MT40A1G16HBA-083E is 1Gb and 16Gb, respectively. The capacity of it is obviously 16Gb. The capacity of memory particles explained in this way may allow us to understand them better.
We compare MT40A1G16HBA-083E to a country with 16 cities, each of which has 1024x1024x1024（1G=1024M，1M=1024K，1K=1024）families. Then this country will have a total of 16x1024x1024x1024 families. If each city has a gate and only one family can be released at a time, this country can only release 16 families at a time.
At present, whether desktop PC or mobile phone has basically entered the 64bit era. The unit of data throughput per processor is 64, that is, the processor needs to draw 64 families at a time. What should we do? So we put together a number of countries, and for a country with 16 cities, it only takes 4 countries to meet the needs of the processor. But if there are only 4 or 8 cities in some small countries, 16 countries or 8 countries need to be united to meet the demand at a time.
Now let's talk about why DRAM or NAND storage particles are not labeled with B but b. In fact, users who know a little about the principles of computers should know that the existing computer architecture B (Byte) represents one byte, while b (bit) represents one bit.
For a single bit of 0 or 1, the recognition computer is "yes" or "no". Computer cannot recognize countless 0 or 1. But how should data correspond to 0 or 1 of the computer? Therefore, this contributes to ACSII encoding. Each letter or symbol corresponds to a ACSII encoding, so that the real world language and the computer is completely docked.
ACSII encoding specifies that each symbol occupies a size of 8bit, referred to as a byte. One byte is the basic unit of storage and the minimum unite of file storage is byte. However, storage unite of DRAM or NAND is not necessarily byte but bit due to the docking of computers or even non-computer devices.
In addition, we will find b as a unit in data flow, such as network bandwidth, USB bandwidth, and PCI-E bandwidth. This is because data transmission is in the form of channel flow, as only one family can be released at a time in the example above. In order to ensure the security of data transmission, some validation data will be added to it, for example USB 3.0 uses 8b/10b encoding method (2bit validation data is needed to transmit 8bit data every time.) At this time, if byte is used as a unit again, it’s obviously out of order and out of date.