The emergence of flash-based storage
February 16, 2010
Right now, it is very likely that there is a spinning disk in your computer. Most desktop hard drives spin at 7200RPM, but some can get up to 15,000RPM. Think about that for a minute. At what RPM does your car switch gears? Maybe 4500 or so?
That quickly-spinning disk has all of your important information on it: pictures, email, documents, maybe even your home business. Desktop hard drive density is now at a maximum of 2 terabytes. That’s 2,000 gigabytes. My first computer had .5mb floppy disks the size of a small dinner plate. Notebook drives currently max out at 500GB, although there are 1- and 1.5TB drives in the works.
Here’s a brief overview of how a hard drive works:
A traditional hard drive has one or many spinning disk platters that vary in data density. They rotate at speeds up to 15,000RPM, and there are little readers called ‘heads’ that sit very close to the spinning platters. By close, I mean tens of nanometers. That’s the width of a few blood cells. Some notebook drives are equipped with motion sensors, to detect when the computer is being dropped, and will ‘park’ the read/write head in a safe place where no damage to data can be done.
SSD stands for Solid State Device. These drives can store large amounts of data, although not at densities that can compete with traditional, spinning hard drives. The SS in the SSD implies that there are no moving parts, so these drives don’t suffer the mechanical faults and bottlenecks that traditional hard drives do, such as power consumption and the possibility of data corruption due to shock.
Some of the characteristics of hard drives are noise, vibration, and data seek delay. The delay is caused by the read/write head having to move around the disk, as well as the time it takes to spin up a hard drive that is ‘asleep’ for power conservation reasons. Samsung, which has pretty much cornered the market on flash memory, along with Toshiba, lays it out pretty nicely in this diagram:
As you can see, a flash-based SSD drive is much different. It uses no moving parts, so there is no danger of a disk head crashing on the surface of a platter. In fact, there are no platters. No spinning. It acts the same way as RAM in a computer uses, except it’s non-volatile, meaning that when you turn off your computer, the contents of a flash-based SSD remain intact. It’s also much lighter, uses much less power, can withstand much more powerful shocks, and operates at a wider range of temperatures. The one difference: Cost. A 500GB 7200RPM laptop hard drive will cost about $200, while an SSD drive would be nearly $2000. More reasonably priced SSD drives are available, but at the cost of capacity.
Being a ‘regular’ hard drive user, I was interested to spend a weekend with a computer that uses a flash-based SSD hard drive. The difference between my computer and my coworker’s were pretty minimal. Mine is an early 2008 MacBook, and hers is a late 2009 MacBook. It has a wider system bus (the “road” between the processor, RAM, and hard drive), and a faster CPU. That’s where the differences stop. I have a high-speed 500GB 7200RPM hard drive in my laptop, while she has a 120GB SSD. The difference was nothing short of incredible:
I hate the sound of my own voice.
Anyway, in conclusion, flash-based mass storage devices carry many advantages over traditional, ‘moving’ hard drives. Like anything new, they cost a lot. And also like anything new, in a year or two the price of these drives will be low enough, at significant enough capacities, to fully replace traditional hard drives altogether. I’ve had a few drives fail on me in the past. One’s motor died. I have years of photos sitting on those platters, inaccessible to anyone without the proper skill to get inside it and fix it. It would also require a clean room to fix.