The application of industrial-grade SSD in embedded applications has great advantages and can gradually replace low-reliability mechanical hard disks (HDD). Today, many embedded systems use traditional storage sizes, including 2.5-inch, M.2 and MSATA, which are small, shock-resistant and drop-resistant, and have high stability, thus improving overall reliability.
For the environment in which embedded systems are located, once a fault occurs, repair is a very difficult task and the cost is not low. With the application of industrial-grade SSDs, the performance of embedded systems has become more prominent. The reason is that it is more durable, has no mechanical moving parts inside, has no mechanical failure, is not afraid of collisions, shocks and vibrations, and can be used in more severe climate conditions for embedded systems.
Typical embedded applications include a small operating system (OS), software applications, and some form of data collection. Applications and workloads for industrial-grade SSDs in embedded systems tend to vary widely. In most applications, the collected data is only temporarily saved for a period of time and then uploaded to a central server for storage. However, some applications are very demanding on the load cycle of the system and require data to be written around the clock, which will affect the use of industrial-grade SSDs. life.
Therefore, environmental conditions are also a consideration. The operating temperature, altitude and humidity of industrial-grade SSD should be based on its own supported parameters. It is important to note that the use of SSD or any storage media will affect the service life of the device and may lead to unnecessary data loss.
Under normal use, the data retention period of industrial-grade SSD flash memory is relatively stable. Different from the magnetic recording principle of mechanical hard disk (HDD), flash memory uses FG (Flat 2D Flash Memory) or CT (BiCS Flash Memory) structure to store electrons and express data content. FG and CT are the keys to continue to store data after the flash memory is powered off.
If the flash memory is stored during a long-term power outage, the electrons in CT or FT are lost slowly, exceeding a certain level, and the stored 0 may be mistakenly recognized as 1, resulting in data errors.
JEDEC (the Solid State Technology Association is the leading standards body in the microelectronics industry) has clear standards for data retention, stipulating that SSDs should store data errors after 1 year (customer grade) or 3 months (enterprise grade) at the end of the flash memory life. Rate. In other words, even if the consumer-grade SSD has exhausted the design life of the flash memory, it can still guarantee that data storage will not fail for one year during a power outage - for flash memory with less wear and tear, this time will be longer.
A similar bottleneck exists in servers because SCSI Attachment SCSI (SAS) and Serial Advanced Technology Attachment (SATA) interfaces cannot keep up with the speeds of industrial-grade SSDs. The new NVMe protocol is faster, and by integrating interrupts and simplifying queue management, system costs can be significantly reduced.