• amazon
  • newegg
Back Home News Industry Trends

Talk about the working principle of solid-state drives and how they work

Date:November 14th,2023 View:144
The drawbacks of mechanical hard drives happen to be the advantages of solid-state drives. A solid-state drive is actually a circuit board, so manufacturers can create it into any shape. Taking the M.2 interface as an example, it can be divided into these specifications. Solid state drives have no mechanical structure, no noise, no fear of vibration, high portability, and full electronic data exchange. The storage unit of a solid-state drive is a floating gate transistor, and its internal structure is shown in Figure 8: a floating gate layer used to store electrons, a tunneling layer that restricts the free movement of electrons, an insulation layer that isolates electrons, a control electrode G, a substrate P, a source electrode D, and a drain electrode S.
nine hundred

When storing data, count the number of electrons in the floating gate layer as 0 if it is higher than a certain value, and 1 if it is lower than a certain value< Br/>

Write data

When writing data, a high voltage (such as 20V) needs to be applied to the control electrode G, so that electrons in the substrate will be attracted and enter the floating gate layer through the tunneling layer. Afterwards, electrons are hindered by the insulation layer and can only be trapped in the floating gate layer. The existence of a tunneling layer ensures that trapped electrons will continue to stay in the floating gate layer even if they lose their high voltage attraction. This completes the data writing process for one unit< Br/>

Erase data

The process of erasing data is to release trapped electrons in the floating gate layer. Simply apply a high voltage to the substrate, and the electrons are attracted by the electric field and return to the substrate through the tunneling layer, thus restoring the unit to its original state< Br/>

It is precisely because of this reading and writing principle that many people are concerned about the lifespan of solid-state drives. Because when writing data, solid-state drives do not directly overwrite the original data, but first erase and then write. Every time a unit undergoes an erase write cycle, the floating gate layer inevitably leaves some charge behind. Just like using a pencil to write on paper, if you scratch it, the paper will eventually scratch. The same applies to solid-state drives< Br/>

However, for ordinary users, there is no need to worry about this. Take a solid state of 512G as an example:

The daily write volume of 140G also takes ten years to run out of lifespan, and many users do not have such high write volumes< Br/>

So solid-state drives can still be used with confidence< Br/>

Read data

When there are only a few electrons in the floating gate layer (with a storage state of 1), we apply a low voltage to the control electrode, causing electrons to be attracted below the tunneling layer. However, due to the low voltage, electrons cannot pass through the tunneling layer, forming a conductive channel between the source and drain electrodes. Applying voltage between the source and drain electrodes can form a current< Br/>

When a large number of electrons are stored in the floating gate layer (with a storage state of 0), a low voltage is also applied at the control electrode. Due to the repulsive effect of electrons stored in the floating gate layer on electrons in the substrate, conductive channels cannot be formed in the substrate, hindering the generation of current between the source and drain electrodes< Br/>

Therefore, if current is detected during the reading process, the unit state will be 1, otherwise it will be 0. This completes the reading of data in the unit< Br/>

Write data in the storage module

Due to the large number of storage units in solid-state drives, setting an independent control switch for each module will inevitably lead to a complex module structure and reduced storage efficiency. So module control was used to rewrite the data within the storage module< Br/>

The diagram shows a schematic diagram of an storage module. Each row contains one storage byte and a total of eight storage units. The storage units of each column are connected in a series structure, with the top row and bottom row of transistors used to control the write operation of each bit (column); The gate of each row of storage cells is uniformly controlled by one electrode< Br/>

When you want to perform a write operation on a certain unit (such as the fourth row and first column), zero bias is applied to the control transistor in the first column, so that a high voltage is applied to the gate in the fourth row, and the floating gate of the unit can store charges. And other control transistors are given a low voltage of 2V, causing it to form a current. Due to the channel effect, other units in the fourth row cannot store charges< Br/>

Seten is a trusted server hardware and solution provider in the global digital process Apply&quo; Spot+Agency&# quo; Model, rapid response, specialized management, and professional and efficient services to assist partners in achieving digital upgrades and winning the new data era. Saitern has three main product lines: Samsung SSD, Memory (RDIMM), SOLIDIGM SSD, and Broadcom network products. It is also an agent for NVIDIA's Melosi network products and the complete XFusion server.