Solid state storage, also called flash storage, has emerged from the fringes and into most devices. Today, the choice of solid state for enterprise storage can radically influence the performance and value you can expect from areas ranging as broadly as backups to data analytics and even user device reliability.
The first flash storage drives appeared in small devices such as phones, MP3 players, and the ubiquitous flash memory stick. Eventually, the technology materialised as solid-state drives (SSDs) that complement or replace magnetic platter drives.
More recently, it took the form of Non-Volatile Memory Express (NVME) drives that are boosting enterprise storage arrays and high-performance computers. All of these storage types share the same characteristics: they are incredibly stable, resilient to damage, and blisteringly fast.
“It’s rare that you have a technology that can be so immediately impactful,” says Malcolm Tiley, technology specialist at Sithabile Technology Services. “For example, if you take an old laptop and upgrade it with an SSD drive, its performance doubles, if not more.
“A boot time that took minutes suddenly happens in seconds. It’s mind-blowing, and very powerful. But flash storage doesn’t fix every storage or performance issue. If you use it in the wrong situations, it can actually cost you more in the long run.”
Types of flash storage
To avoid such problems, it’s good to understand the different types of flash form factors. First, how does flash storage work? Magnetic storage, like a hard drive, manipulates magnetic fields to record information. Optical storage uses light to write and read data from a special surface, such as a DVD. Flash storage uses grids of electrons and changes their state with electrical currents to record data.
This approach has two distinct benefits: it has no moving parts that could be damaged through hard knocks or drops and doesn’t require special equipment such as a DVD drive to read the data. The manipulation of electrons is also very fast: often four to five times faster than other storage types. Another advantage is power: flash systems require less energy and don’t need time to ‘spin up’, making them immediately operationally available.
Three common form factors use flash storage:
* Flash thumb drives: A thumb drive plugs into a USB port. Useful for file transfers and keeping data with you, flash drives are generally as fast as the USB port they use and the quality of their electronics.
* Solid-state hard drives: SSDs are interchangeable with magnetic hard drives and use the same plugs and cables. They are swift and often used as primary drives, though they can become expensive for large storage capacities, such as two terabytes and up.
* Non-Volatile Memory Express: NVME drives are long, narrow cards designed to interface more directly with computer hardware. Though very fast, they require a particular hardware port, and their performance is often faster than what most people need.
Best use cases
Generally speaking, flash storage is more expensive per gigabyte than other storage mediums and offers less capacity. But they compensate with incredible resiliency and speed.
The most visible use case is SSDs in laptops and as external hard drives. They are much less prone to fail, making them ideal for the movements and knocks of a mobile device or a hard drive that you carry around with you. That being said, SSDs can fail and have a lifetime limit, though such a drive will last for many years under normal operating conditions.
SSDs are expensive in terms of the storage you get. It’s often cheaper to buy a large storage magnetic hard drive of 3 terabytes or more than to get an equivalent SSD. Many high-end PCs use an SSD for its primary drive to ensure fast boot times and application performance, and add a magnetic drive for extended storage.
Flash arrays are storage systems that house multiple SSDs. These systems are similar to standard storage arrays that use redundancy measures such as RAID (writing data across multiple drives to make recovery simple if one drive is lost).
But they are much faster and employ special storage controllers and compression techniques for even better performance and space optimisation. Such arrays are often used for high-speed data transactions.
NVMEs are primarily used to create high-end flash arrays. On paper, they are far superior to SSD arrays, though in practice, not every use case requires such incredible speed. NVME arrays are usually reserved for very demanding data systems, such as large cloud platforms and high-end gaming and AI systems.
Which storage is best?
Replacing all your storage with SSDs or even NVMEs is tempting. But that might be a mistake, says Tiley: “The best storage depends on the situation. If you have archive data you rarely access, then magnetic tapes of optical discs are perfectly fine and very affordable.
“When storing large pieces of data, magnetic hard drives are still the leaders in terms of capacity and cost. Flash arrays are very fast, but most organisations don’t need them, or they already use them indirectly when they consume cloud services.
“On the other hand, if you use a regular storage array and you keep running into performance bottlenecks, it’s time to upgrade.”
Flash storage offers more choices for companies as they optimise their backups, data streaming, and storage capacity. One size does not fit all, but the right fit will make a huge difference.