In this blog post, we will examine how SSDs surpassed traditional HDDs and take a detailed look at the technological advances and background behind their establishment as the next-generation storage medium.
If you’ve recently tried to buy a computer, you’ve probably heard the advice to “make sure to get a computer with an SSD.” Before SSDs became mainstream, hard disk drives (HDDs) were the dominant storage medium, praised for their speed and storage capacity. HDDs, first developed by IBM in 1956, have evolved steadily over the decades and became the standard for data storage devices.
During this time, many companies and individual users relied on HDDs to efficiently store large amounts of data. However, as time passed, the limitations of this technology became apparent, and the need for a new alternative arose. Despite the performance of hard disks, SSDs, which have been commercially available for less than 10 years, have established themselves as the next-generation storage medium with speeds dozens of times faster than hard disks.
Unlike hard disks, whose performance has plateaued, SSD performance is expected to continue improving in the future. The invention and adoption of SSDs have brought about tremendous changes, particularly in the IT industry. For example, data centers that have adopted SSDs can now support more customers effectively thanks to significantly faster data processing speeds. So how did SSDs achieve such high growth potential?
An SSD (Solid State Drive) is primarily composed of three components: an interface, memory, and a controller. The interface is responsible for transferring data to and from RAM. While traditional HDDs did not require high-speed interfaces due to their internal performance limitations, the explosive speed of SSDs demands more advanced interfaces. There are several types of interfaces, including SATA, SATA Express (faster than SATA), and M.2 (used primarily in devices like laptops due to its small size). These interfaces play a crucial role in determining the data transfer speed between the computer and the SSD, thereby influencing the overall performance of the SSD. In particular, the introduction of the NVMe (Non-Volatile Memory Express) protocol has enabled SSDs to achieve significantly faster data transfer speeds compared to previous-generation interfaces.
Memory is where data received from the controller is stored or sent to the controller. Before the commercialization of SSDs, there were products that adopted RAM-based memory. However, due to the characteristic of RAM that data is lost when power is cut off, SSDs are now only manufactured with flash memory-based memory. Flash memory is non-volatile memory, which has the advantage of retaining data even when power is turned off. Memory types are divided into SLC (Single Level Cell), MLC (Multi Level Cell), and TLC (Triple Level Cell). Memory is created by connecting hundreds of millions of small units called cells. In the case of SLC, a single cell can store 1 bit of information, either ‘on’ or ‘off’. In MLC, a single cell can store 2 bits of information, and in TLC, a single cell can store 3 bits of information. Recently, a new type of memory called QLC (Quad-Level Cell) has emerged, which can store even more data. However, as more data is stored in a cell, the cell’s lifespan decreases, and the speed of reading and writing data tends to slow down. To address these issues, SSD manufacturers are researching various lifespan extension technologies and performance optimization methods.
The controller connects the memory to the interface. An SSD controller requires much more complex technology than that of an HDD. Due to the limited lifespan of flash memory, it is essential to have technology that prevents specific cells from being repeatedly used. Additionally, the controller must remember the addresses where specific data is stored and retrieve that data. Creating an efficient algorithm to identify these addresses is a challenging task, so the performance of the controller directly affects the performance of the SSD. Recently, controller optimization using AI technology has been studied, and it is expected that this will further improve the lifespan and performance of SSDs. Furthermore, the address storage method of SSDs is fundamentally different from that of HDDs. As a result, older operating systems (products prior to Windows Vista) recognize SSDs as HDDs, leading to inefficient data management. Therefore, installing an SSD in a computer running an operating system like Windows XP is not recommended.
What are the advantages of SSDs over HDDs? First, they generate less heat, resulting in a higher operating temperature limit than HDDs. HDDs use a motor in the center to spin thin metal plates, and a head reads data from the plates using light. Increasing the motor’s RPM can cause the motor to overload, damaging the metal plates. Similarly, reducing the wavelength of the light from the head to decrease the physical size of the data can overload the light source and reduce durability. While SSDs also generate significant heat when operating at 100% performance, such conditions do not occur unless artificially created. Therefore, SSDs have the potential to achieve higher clock speeds and higher cell densities. Lower heat generation enhances system stability and serves as a significant advantage for users requiring high-performance tasks.
Second, SSDs can also have significantly higher storage capacity. Commercially available 3.5-inch HDDs can have a maximum capacity of around 20TB. HDDs can be produced in sizes larger than 3.5 inches. However, this means increasing the size of the metal platters, which in turn reduces the motor’s RPM, leading to decreased performance and making practical implementation challenging. In contrast, as SSD technology advances, it is becoming possible to produce high-capacity SSDs. As a result, SSDs have gained an advantage over HDDs in terms of high-capacity storage, which was previously a strength of HDDs. Recently, SSDs with capacities of tens of terabytes (TB) have begun to be commercialized, and their adoption is increasing in large-scale data centers such as cloud storage facilities. As a result, SSDs have become an important storage medium not only for personal computers but also for enterprise data centers.
The only remaining advantage of HDDs is their lower cost per unit of storage capacity. HDDs still offer a cost-effective solution for storing large amounts of data, which is advantageous for large-scale backup systems or high-capacity file servers. However, SSD prices continue to decline, and cost-effectiveness is improving alongside technological advancements. While both SSDs and HDDs will coexist for the foreseeable future, it is anticipated that SSDs will become the more prevalent storage medium in the coming years.
