All about SSDs
Solid State Disks (SSDs) come in a variety of shapes and sizes and they work in a totally different way to a conventional hard disk. Whether you're replacing a failed SSD or upgrading from a hard disk, these are things you may need to know about.
Summary
This page is in two main sections: firstly we discuss the different shapes, sizes and types of SSD, and then we explain how they work and hence how they need to be treated differently from a hard disk.
But first let's get one thing straight: a Solid State Disk is no more solid than a hard disk, and it isn't even disk-shaped! (The only liquid state computer memory that has ever been used was the mercury delay line, and those went out of fashion in the 1950's!) An SSD would be better called semiconductor, electrostatic, non-magneting or non-rotating storage, but we're stuck with the name we've got.
Safety
- Consider whether your page should include a specific section on safety.
Shapes and Sizes
The meat of the page goes here, typically organised under several top-level headings.
Under the Bonnet
Unlike a hard disk, an SSD contains no moving parts or delicate mechanisms and so is much more rugged and tolerant of physical shock. But that's only the beginning.
The way in which an SSD works is radically different from a hard disk is several respects:
- An SSD stores data as electrostatic charges on a silicon chip whereas a hard disk stores it as patterns of magnetism in a spinning magnetic disk.
- When you write data to a hard disk it simply overwrites any previous data in the same location whereas you can't write zeros to an SSD, you can only write ones (all data consists of ones and zeros). Hence the area you want to write to has to be erased (set to all zeros) before you start writing.
- You can't erase a little bit of an SSD here and a little bit there. It only allows you to erase quite large block of memory at a time.
- You can write to a hard disk as many times as you like without wearing out the magnetic surface, but writing to an SSD causes wear. An SSD will therefore choose where to put your data in order to spread out the wear and will itself remember where it put it. This is known as wear-levelling.
A consequence of all that is that as you create, edit and delete files, more and more blocks of storage will contain some data which is still current and quite a lot which is not (deleted and old versions of files, known as "stale" data). Eventually you would end up with no empty blocks to write new data to even though the SSD might only be half full of current data. To deal with this problem the operating system (Windows, MacOS or Linux) tells the SSD which data is no longer needed through a process known as TRIM, and as a background task, the SSD tries consolidate the current data into fewer blocks, freeing up blocks which can then be erased and re-used. This process is known as garbage collection.
All this represents a lot of work for an SSD to keep track of where it's put all your data and which bits of it are still needed and which aren't. So if anything goes wrong it can go very badly wrong, leaving the SSD totally confused and no longer able to operate. This might happen, for example, if power is suddenly lost or a power surge occurs while the SSD was in the middle of doing garbage collection, even though SSDs (certainly the better ones) are deigned to cope with such an eventuality. By contrast, when a hard disk fails it often does so more gracefully with increasingly frequent read errors, giving you more chance to recover your data. They may therefore be more suitable than SSDs for backup.
External links
- External links (if any) as bullet points.
- If non, delete this section.