The basic working principle of a mechanical hard disk (HDD).ssd drives
When a mechanical hard drive needs to read or write data, it receives a command, the head moves to the appropriate position, and the platter rotates so that the data operation area is in the specified position.
The time required for these actions is the seek time and latency period, which take several milliseconds due to the need to move the device.
This is due to the read/write mechanism of the operating system:
Hard disk is divided into several areas as the most basic operating unit, the unit is called "sectors", when a new writing data, choose one or a few sectors for writing data, the position of these sectors are next to each other, they are logically continuous, in both the read and write the required time is short.
And here's the crux of the matter:
After all the data are not in the established forever there won't be changed, when previously written data modification, such as increasing the content and data volume increase, and next to the location of the original sectors have other data, these new data will be written to the other place to go, so we see in the operating system of a file, in the actual physical address is not continuous,
When the file is read again, the disk will have to do more work, and in the worst case, the head and platter will be moved and rotated many times, and the final working time will be doubled.
With the increase of storage capacity per unit area, we can enjoy higher capacity hard drives, but the speed of reading and writing data has not been much improved.
Because the key factors that determine seek time and latency period: head movement speed and disk rotation speed are close to the limit, further increase will bring other adverse factors: such as cost increase, noise, temperature increase and so on
1.Physical structure of the ssd drives
The system interface, power supply, and drive mode of SSDS are all the same as those of HDDS. The main changes are the constituent units and physical working mode.
The SSD internals include a PCB board, a main controller chip, a flash memory chip and, in some cases, a cache.
The most basic unit of an SSD is a Flash memory chip, called Nand Flash. It is a non-volatile memory chip that can write and erase data very fast by charging and dropping data.
Since signals are transmitted entirely through circuits during read and write operations, there is no movement of the head, rotation of the platter, etc., as in HDDS, which greatly reduces the processing time.
However, Nand Flash is also divided into several types. Currently, consumer SSDS and even many enterprise SSDS use MLC (multi-layer cell) Flash memory, which has a lower write performance than SLC (single-layer cell) Flash memory, and a shorter life span, but the price is much lower.
Even so, the current cost of SSD has not been reduced to the extent that everyone can accept, and the price is still the barrier affecting the further popularity of SSD.
An SSD is composed of multiple Nand flash flash memory particles, each of which can be considered as an independent storage unit, and the master controller makes a RAID parallel between them.
This means that SSD reads and writes are "multithreaded", and the work is not limited to one particle at a time. The master can split the data and write it to different particles at the same time, which is naturally faster.
This is one of the reasons why SSDS are fast.
Of course, what the master has to do is fa
2.The master controller of the ssd drives
SSD Flash memory also has a minimum operation unit. Compared with mechanical hard disks, Nand Flash has a special difference in that the minimum unit for write and erase operations is different. The minimum unit for write operations is 4KB, which is called "Page", while the minimum unit for erase operations is 512KB, which is called "Block".
That is, writing to a blank cell can be done on a page-by-page basis, but to delete the data, the entire block needs to be erased.
When the data in a block needs to be deleted, the data to be deleted will be marked rather than physically erased. Then when the data needs to be written in the same physical location again, the valid data will be retained, copied to the new block, and then the original block will be erased.
It sounds complicated, but to put it simply, the SSD write mechanism requires 1MB of data to be written. The actual operation volume may be larger than this. The specific amount depends on whether the main controller's algorithm is efficient, and the actual random write speed depends on whether the operation speed is fast enough.
Like HDDS, SSDS require logical addresses to be managed. However, the minimum operating system logical address unit is 512B, and the minimum SSD write unit is 4KB, which requires that, the chipset, and the main controller work in sequence.
In addition, the main controller and is responsible for the allocation of tasks, each flash chips all flash memory state monitoring, the management of each piece of data validation, etc., quite a lot of work and multifarious, and this is why in some of the new master ARM to use dual core processor, because master performance will directly affect the speed of the SSD.
3.Advantages of ssd drives
Fast read/write speed: The flash memory is used as the storage medium, and the read speed is higher than that of a mechanical hard disk.
Solid state drives are not used, and the seek time is almost zero.
The sustained write speed is amazing, and most SSD manufacturers will claim that their SSD sustained read/write speeds exceed 500/s!
Solid state drives are faster than just continuous read and write. Random read and write speed is the ultimate meaning of solid state drives, which is most directly reflected in most daily operations.
Also associated with this are extremely low access times, with the most common 7200RPM mechanical hard drives typically having seek times of 12-14, while solid state drives can easily reach 0.1ms or less.
Shockproof and smash resistant: Traditional hard disks are disk-shaped, and data is stored in the disk sector.
And solid-state drives is to use flash particles (i.e., mp3, U disk, etc), so there are no mechanical components inside SSD solid state hard drive, so even in the case of high-speed mobile even with flip tilt also won't affect the normal use, and in the case of collisions and shocks to minimize the possibility of loss.
Compared with traditional hard disks, solid state drives have absolute advantages.
Low power consumption: Solid state drives are lower than traditional hard disks.
No noise: Solid state drives have no mechanical motors and fans, and the noise level is 0 dB when working.
Flash-based solid-state drives consume less energy and generate less heat when operating (although higher-end or high-capacity devices consume more energy).
There are no mechanical moving parts inside, no mechanical failure, and not afraid of collision, impact, vibration.
Since solid-state drives use flash memory chips without mechanical components, they have the characteristics of low heat and fast heat dissipation.
Wide range of operating temperatures: Typical hard drives only work in the range of 5 to 55 degrees Celsius.
Most solid-state drives work from -10 to 70 degrees Celsius.
Solid-state disks are smaller and lighter than mechanical disks of the same capacity.
The interface specifications, definitions, functions, and usage methods of solid state drives are the same as those of common hard disks, and the product shape and size are the same as those of common hard disks.
Its operating temperature range is very wide (-40~85 degrees Celsius).
Lightweight: Solid-state drives are lighter in terms of weight, weighing 20-30 grams less than regular 1.8-inch hard drives.