What is the market advantage of HDD compared to solid state drives?

Last Update Time: 2020-09-19 10:49:33

With the rise of solid state drives (SSDs), people seem to be slowly forgetting about mechanical hard drives (HDDs), the storage "partners" that have been with us for more than half a century. Who remembers the first mechanical hard disk? In 1956, IBM introduced the first mechanical hard disk (HDD). At the time, the mechanical hard disk was larger than a domestic refrigerator, but it could only store 3.75MB of data. But at the time, it was a revolutionary technology.

With the development of technology, the volume of mechanical hard disks is getting smaller and smaller, the speed is faster, and the data that can be stored has reached terabytes or more. But in these 60 years, HDD technology has not changed revolutionarily. HDD and tape have always dominated the desktop and data center storage world. Until nearly a decade or so, NAND flash solid-state disks began to mature, so much that they rivaled or even surpassed HDD in terms of capacity. lower and lower.

Most experts believe that solid-state drives are destined to replace HDDs as the primary storage medium. But judging from today, there is no direct answer to the choice between SSD and HDD, and this is a long-term topic. Users need to consider this issue based on multiple factors such as application scenarios, latency, bandwidth, and storage architecture. Today we will talk about the choice of SSD and HDD. Traditional HDD stores data on a high-speed rotating disk. When the disk rotates, a pair of magnetic heads moves on the disk to read or write data.

Data is stored on circular tracks, which we call tracks. Each track is divided into a number of small areas, which we call sectors. Most disks use a pair of platters, which are mounted on a central spindle with a small gap between them. Unlike HDDs, SSDs have no moving parts. Instead, SSDs write and store data on flash memory chips. SSD manufacturers stack storage chips in a grid to achieve different densities. To prevent fluctuations, SSDs use floating-gate transistors to store charge. This technology allows the SSD to retain the stored data even when no power is connected.

For customers, there are currently several different types of SSDs available on the market: SLC: The SSD stores one bit in each unit. This design improves durability, accuracy, and performance. For enterprise critical applications and storage services, SLC is the flash technology of choice. Of course, it has the highest price. TLC: Each unit stores 3Bit, which is usually used for consumer electronics with relatively low performance and durability requirements. Best suited for applications with a large number of read operations, and has a place in read-intensive enterprise storage applications.

eMLC: is an "enhanced" version of MLC NAND flash, which to some extent bridges the performance and durability gap between SLC and MLC. eMLC drives are more expensive than MLC drives, but much cheaper than SLC drives. Although each unit still stores 2 bits, the controller of the eMLC drive manages data placement, wear leveling, and some other storage operations to extend the life of the eMLC SSD.


image.png


MLC: A multi-level unit SSD that stores 2Bit per unit, which is usually considered as a consumer technology. Although filling 2Bit of data into a storage unit can save space, the trade-off is short life and reduced reliability. MLC SSDs are commonly used in desktop and notebook computers.

QLC: 4 bits of data can be stored in each flash memory cell, which provides more capacity than SLC, MLC and TLC, but the reliability may not be good. Nonetheless, QLC's cost, density, speed, and power efficiency attributes make it ideal for applications such as machine learning, data analysis, and media streaming. Unlike HDDs, all types of SSDs are "consumables," meaning they wear out as data is repeatedly written to the drive. Failures of solid state drives usually occur gradually. With the failure of a single unit, the overall performance will gradually decrease.

Many SSD manufacturers address this kind of emerging problem called "wear and tear" by over-provisioning their products. That is, the actual capacity is more than stated in the product brochure. But companies don't have to worry about this situation, as long as the workload meets the DWPD rating, SSDs can easily be used throughout the warranty period. At present, most enterprise users use SSDs with TLC technology, mainly because they cost less than other types of flash SSDs. TLC SSDs are usually used for regular read tasks and light write operations.

QLC SSDs are characterized by low DWPD and have advantages in density, speed, and power efficiency, so they are often used in high-performance, read-intensive applications. At the same time, more and more IT organizations seeking higher performance are turning to 3D XPoint-based SSDs, which are a new class of non-volatile storage and storage devices that are faster and denser than previous NAND flash devices.

Generally speaking, the performance of SSD is better than HDD. SSD has higher performance, reliability and energy efficiency. In addition to reducing the energy cost of data centers, SSDs can also reduce the space occupied by data centers because more space can be used. Fewer SSDs produce the same IOPS as hard drives. However, SSD is not always the best choice for every enterprise data storage. The main determinant is still very much related to your application scenario. For example, due to high cost and low usage, SSD is rarely the choice of old file storage requirements.

Cost is always a key factor in determining HDD or SSD in storage tasks, especially since the current price of SSD is four to five times that of similar HDDs. Therefore, when choosing, you need to comprehensively choose whether to use HDD or SSD. Fortunately, the emergence of new technologies such as 3D XPoint has provided better cost performance.

In addition to cost and long-term wear issues, the biggest problem associated with SSDs is that SSDs often fail directly without warning. However, before the failure of traditional mechanical hard disks, there will usually be a warning that the performance is lower than normal, which will allow users to provide early warning. The SSD needs to be properly monitored with software, otherwise sudden failures will crash enterprise users.

The best way to start planning the deployment of modern storage systems is to first determine the choice of storage media according to your application scenario. If you frequently deliver a large number of files to many users, the more SSDs you run, the better. On the other hand, if you only need a storage system that uses old files occasionally, you should focus on using traditional HDDs.

However, price is still a very important factor for customers to choose. If the price of SSDs has dropped significantly in the next few years, then the hard disks of many data centers will soon be migrated to SSDs, and HDDs may permanently exit the data center . Generally, the only reason for not migrating from HDD to SSD is cost. Considering all the advantages of SSD, the price is becoming less and less problematic.

 

If you want to know more, our website has product specifications for the solid state drives, you can go to ALLICDATA ELECTRONICS LIMITED to get more information