The research department at Technospares was having lunch together when the topic of Solid-State Hard Drives (SSD) drives came up. It was interesting to discover we were all moaning about the lack of information and misinformation around this topic. So we want to set the record straight. While we are going to be discussing SSD’s in general the information applies to both internal and external along with internals being used in hard drive caddies.
The first thing we want you know is not all SSD’s are created equal. First there is one of two types of storage for the flash memory. One is known as single cell while the second is known as multi-cell. Not long ago multi-cell was considered a poor performer and did not have the durability for server use. However single cell SSD’s such as from Intel were viewed as real budget strainers and suitable only for when the ultimate in performance was required.
Multi-cell has come a long way in the year 2012 and is now capable of enough write changes to where in average use a life expectancy exceeds five years. This highlights a fundamental difference from a mechanical hard drive. The magnetism in a traditional hard drive does not wear out. The circuit boards or bearings are more likely to fail.
Solid State Hard drives have a moving parts count of exactly 0. This makes them much more impact and shock resistant. They also draw less power by at least an order of magnitude. The downside is changing a zero to a one (on any type of flash memory (from USB stick to SSD drive) you have consumed a state change which has a finite count, not unlike a biological heartbeat.
High heat can reduce the total count of these changes that are available. This heat can come from changing the bit on and off in a local area of the flash memory quickly over a period of time. It can create a hotspot that will burn out that segment, or the entire storage device.
For this reason you absolutely do not want to employ any sort of defragmentation utility. Your data is not scattered across a large surface area as in a mechanical hard drive.
The manufacturers’ war in SSD drives comes not from a wide variety of the quality of flash memory. It comes from the controlling chip and its performance. This is where the differences lie. This chip does all the management and that includes what is known as wear leveling. The idea is to use fancy algorithms to distribute the state changes that occur evenly over the entire device.
SSD wear levelling actually becomes very complex and quickly. This is because in Solid-State Hard Drives (SSD) storage data is stacked more like towers of coins. To change the value in the middle of the stack the entire stack has to be moved to a different stack, changing the one coin as it is moved.
And that is just the beginning. This fancy controller mentioned above attest to minimize the amount of stack changes to minimize wear. In essence it creates temp files. Later during an idle time (this could be measured in seconds), the controller chip will run a sort of a garbage collection process to clean up the storage area.
This last topic becomes complex in itself depending on the controller and the operating system.
The last thing you need to know is to watch out for how thick a drive is. Replacing a drive in an older laptop can be very cost-effective for improving performance and extending its somewhat used battery. The issue can be your SSD drive could actually be too thin to fit height wise in your laptop. You may need to plan on an adapter bracket or get creative with some padding.
Generally speaking most of our clients would find any of our SSD drives suitable for anything outside of enterprise class server storage. You can see a selection of the SSD we have at Technospares here. Providing a list of amazing products and couple it with even better customer service.
Leave your comments below, if you liked reading Solid-State Hard Drive (SSD) Myths, Tips and Tricks