This article is intended for non-specialists wanting to know a little bit more about SSDs.
The solid-state drive (SSD) is a relatively recent addition to the technologies available for mass data storage. In place of the spinning magnetic disk used in hard disk drives (HDDs) since the 1950s, an SSD relies on solid-state digital chips to store information. In recent years, SSDs have seen increasing use in many computer systems, from laptops to commercial web servers. Although SSDs offer clear benefits such as faster performance, the technology has a few limitations worth considering.
The marketplace currently offers a few different SSD memory chip technologies, each intended to fill specific needs. Among these is a low-cost NAND flash device called Multiple Level Cell (MLC) that gives the most bytes for the dollar, and performs well enough for consumer use. At the high end is Single Level Cell (SLC), which costs more but is faster and has longer device life.
SSDs have no moving parts and can store and retrieve data faster than a traditional hard disk drive. In an HDD, a mechanism scans the data recorded on the surface of a spinning metal platter. Due to physical inertia, it takes a few thousandths of a second to locate and fetch information. Although this seems quick, solid-state memory, not having that physical inertia to deal with, can perform much faster. In general, an SSD will outperform an HDD by up to a factor of 1,000, with random reads/writes racking up the biggest improvement, and sequential writes showing the least. In addition to faster speed, the lack of a motor-driven mechanism means the SSD is completely silent. SSDs are also more rugged than their mechanical cousins, standing up better to everyday bumps and jolts. Most SSDs also consume less power and physical space than HDDs, so they are a growing and popular choice in storage for laptops (not to mention tablets).
Wear and Tear
Although SSDs have no mechanical moving parts, each memory cell degrades electrically when writing new data over old. This means the drive can read data indefinitely but writing takes its toll on the memory chips. Depending on the specific SSD technology, any given memory bit can be rewritten from 5,000 to 100,000 times. When a bit is degraded, it can no longer reliably hold data. At this point, the drive’s controller circuit automatically moves the data from it and neighboring memory cells to a “fresher” area and marks the worn area as “out of service.” The controller skips the marked area for all future use.
Because of the memory wear issue noted above, most modern SSDs now come with wear-leveling technologies that keep track of where data is written. This technology avoids repeatedly over-writing the same physical bits and spreads the wear throughout the drive. Wear-leveling prolongs drive life and postpones the time at which the drive takes blocks out of service.
SSD manufacturers build extra capacity into each drive, amounting to roughly 7 to 30 percent of the rated capacity. This practice, called over-provisioning, ensures that the drive maintains its rated capacity for the reasonable operating lifetime of the drive, despite losses from worn bits. The extra room also helps maintain drive performance as the drive fills with data. In addition to factory over-provisioning, you can manually adjust the drive’s overhead space with utility software.
As of 2016, SSDs come at a premium price, offering less capacity for each dollar spent compared to HDDs. The storage capacity of a given SSD also tends to be somewhat less than HDDs, so if you manage many terabytes of data, you’d need more SSDs than you would the cheaper and larger HDDs.
Formats and Compatibility
Many consumer-grade SSDs are available in the 2.5-inch standard drive format used in laptop PCs. Other formats include mSATA and the newer M.2; these slim, card-style designs use a fraction of the space of traditional drives, attaching directly to the motherboard through PCIe or a dedicated socket. Manufacturers make available adapter brackets that allow you to fit these new form factors into the older 3.5-inch HDD slot or into interfaces that might not exist on some older motherboards.
Consumer vs. Enterprise Grade
Solid-state drives come in consumer or enterprise-grade units. Consumer-grade drives tend to be less expensive but still well-suited to everyday personal tasks. Enterprise-grade drives tend to be faster and more expensive with enhanced, brownout-resistant power supplies and memory chips that hold up better under continuous, write-heavy workloads.
SSD Economic Trends
Since the 1960s, solid-state memory chips have increased over a millionfold in capacity. This dramatic trend will likely continue as semiconductor makers push the limits of their science. The price of SSDs has steadily fallen over the past several years, and capacities have risen. It seems all but inevitable that SSDs will overtake HDDs at some point, rendering them obsolete.
Solid-state drives represent serious competition to traditional HDDs. Although currently more expensive and not without their own technical issues, they are clearly faster than mechanical hard drives, and as the technology advances, the benefits will only improve with time.