The evolution of Computer Storage (Hard Disk, SSD and more)
How compute persists billions of bits of data?
Ever wondered, how your computer has prodigious memory storage in which it can store billions of bits of information?
The need to store information on a computer has been around ever since there are computational devices of any sort. Scientists and engineers have been working relentlessly for years to make the computer storage more compact, fast, reliable, power-friendly, lighter and cheaper. With every passing decade, new technologies have revolutionized the market, and the latest one to join the list is the Flash memory which is better known by its famous application the Solid State Disks (or SSDs).
But before we deep dive into what that is all about, let me take you to a journey of how computer storage has evolved over the years.
Using Papers (Punch Cards)
The first data storage of any sort that the computers used in its history were the Punch Cards. It is a piece of stiff paper that can be used to contain digital data represented by the presence or absence of holes in predefined positions.
Paper was cheap and mostly durable, and it didn’t need power supply. So, they quickly gained popularity to store data. The punch card designed by IBM was of the dimension 80 columns and 12 rows storing 960 bits of data. That fueled the culture of having 80 characters as a ‘standard’ for code width in many terminals and code editors.
But they were ‘write-only-once’, ‘non-refresh-able’ form of memory since you cannot un-punch a card. So, the need for new form of computer storage was heavily due.
Using Magnetism (Magnetic tapes, Hard disk drive)
It is based on the principle that certain metals can be magnetized (like Iron) when they come in contact with a magnet. They retain their magnetic property even when the magnet was removed from their presence. Their magnetized state can be used to store information.
What’s more interesting, the magnetization direction (aka Polarity) of the metal can be reversed. The computers uses this property to store information. One direction can be used to represent 1 and the other as 0.
If you happen to open up your computer to peek its internal parts, I’ll bet you would never find a physical magnet. Instead, your computer uses an electromagnet (the kind of magnet that can be formed by passing an electric current through a coil of wire).
When the power is on, the computer can use that electric power to change the magnetic storage (changing 1’s to 0’s and vice versa). Even when the power goes off, the magnetic storage can retain its state (thus providing the persistent storage).
Remember the old cassettes which stored your favorite 90’s songs? (I’m appealing to only 90’s kids !!)
It used the above principle to store data. It is formally called as magnetic tapes. Early computers also used these tapes to store persistent data.
The surface of the tape is made up of a magnetic material which is divided into many tiny sections. Each tiny section can be magnetized independently to store binary information.
Remember the cumbersome rewinding or forwarding these tapes to re-run your favorite song? It had to be done so, because the data in the tape can only be accessed sequentially, and not randomly.
To read or write binary data to a particular section of the tape, that section has to be placed under the read and write head. For which we will have to roll the tape to that point (Which is way too slow!).
There you have it lads! In comes the Hard Disk Drives (HDD) which we all are so familiar with. It is based on the same core magnetic principles to store information but it provides random access to any memory location. (Pretty Cool!).
Instead of a tape it uses a disk (also called a platter) whose surface is made up of a magnetic material.
The read-write head moves back and forth to access different circular tracks of the disk.
The disk rotates very quickly (usually 10,000 RPMs) to allow the read and write head to access specific section (called sector) in that circular track.
Typical HDD has many such platter stacked over each other, and each having two read and write heads to write on either side of the platter. This makes lot of data storage possible in a very compact space.
Using light (Optics) (CD, DVD, Blu-ray)
When you throw a beam of light, a smooth surface will reflect it straight back while a bumpy surface will bounce it off somewhere else. It is this property of light that can be used to store binary information.
Most of the early CDs and DVDs were only single-time writable devices. Once the information gets ‘burned’ on them, it cannot be changed. But why so? And why is it called ‘burned’? Well, the answer lies in the technique called ‘Optical Laser Technology’ which is used to write data on the disk.
On the shining surface of the disk, a laser beam is shot to burn the surface and create a small microscopic bumps called (pits). They are referred as 1’s.
The other un-burned surfaces are called lands. They are referred as 0’s.
In this way, during the writing process, the laser beam can burn the binary data on the surface of the disk. The reader head (the CD, DVD player) shoots light on this surface, and the way it is reflected back will determine whether the surface was 0 or 1.
There is another curious property of light. When it is thrown on substances, some of them reflect it back, while some of them absorbs it. Based on this, another alternate technique to burn the CD was introduced. It didn’t create lands and pits, instead it creates patches of light absorbing or light reflecting areas. Those that reflect light becomes 1, and those that doesn’t become 0.
So far so good. But as time went on, these ‘write-only-once’ disks weren’t cutting the deal. They were well suited to the use cases of distributing music and software, but not for storing user data which requires multiple writes. So, a new kind of technology arose. The writable disks.
To understand this, we need to understand atoms. As we know, the atoms arrange themselves in a different patterns in solid, liquid and gases. In solids they are tightly locked together, in gases they are free to roam around. Some kind of atoms (or molecules) can arrange themselves in multiple different ways even in the Solid States. They are called Solid Phases.
What’s more interesting is that, there are some solid materials which can move back and forth between these different states. These are called Phase Change or Phase Shift.
In some solid state, they allow light to pass through (called Crystalline state) which is referred as 1, and in some state they absorb light (called Amorphous state) which is referred as 0. It is this property that can be used to store binary information. These solid states are not permanent and can be toggled at will, allowing us to write the data as many times as we want.
The DVD and Blu-ray works on the same principles as CD. The only difference is the laser wavelength which is used to create and detect pits. DVD uses lower wavelength red laser beam so it can have lower pit size than CD, allowing more data to be stored in almost the same size. Blu-way disks uses Blue laser light which is even lower wavelength then DVD, so even lower pit size, thus more data storage. The blue laser beam gives it the name Blu-ray.
Both magnetic and optic technologies were handy to store binary information, allowed random access to data, durable, and cheap but their major drawback was the moving part, the mechanical read-write head, which had to be navigated to read-write data, back and forth. The race was on to find the new speedy technology. In comes the Flash Memory.
Using electricity (Memory Cards, USB drives, SSD)
After all the computer is an electronic device, so, an electric solution to storing information was bound to arise. It is all down to semi conductors and transistors (more specifically field effect transistors). The physics behind it is simple enough to understand.
It basically has three components. A source (through which the electric current originates), a drain (through which the current moves out on the other side), and a gate which can allow or block the current to pass through.
When the gate is on and the current is flowing, the transistor is on (this can be referred as 1). When the gate is off, the current is not flowing, the transistor is off (this can be referred as 0). This way we can represent binary information in a transistor. But the drawback of this kind of transistor is, it gets switched on or off using electricity.
But, when there is no power supply, all the gates and source gets switched off too. There is no electricity moving through any of the transistors. When the power supply is back on, there is no way to remember what was the previous state of the transistor. So, it forgets things.
A flash transistor is different because it has a second gate (called floating gate) just below the first one (the control gate). When the control gate opens, some electricity leaks up in the floating gate (through a process called Quantum tunneling.) and stays there. Because the floating gate is completely surrounded by an insulator, the electrons gets trapped there, recording a number one.
Even if the power is turned off, the electrons will stay trapped there between the two gates for years. That’s how the transistor stores its information whether the power is on or off. The information can be erased by making the “trapped electricity” drain back down again (again through the genius of Quantum Mechanics).
It is called Flash memory because it can read or write data in a Flash. The Flash memory has been providing persistent storage to camera, mobile phones, USB drives, memory cards and the latest sensation is SSD (the Solid State drives).
The SSD is built on the flash memory technology. It is solely built using microchips, and contains no moving parts. It mainly contains two important parts,
- Multiple Flash Memory Chips: These chips contain billions of Field Transistors. Each of them can store data without any power supply, as explained above.
- Controller: It takes instructions from the Computer CPU, and directs the read/write of data in the flash memory chips. The read operation happens by passing electrical signals to the flash memory to search for trapped electrons. Write operations happens by sending electoral signals to drain the trapped electrons. Since, there is no physical moving part to read or write data, the speed of read/write is enormously faster compared to HDD.
- Cache: The SSD also contains Cache to optimize the read time further for frequently accessed data.
As we saw, the SSD gives us faster access time, consumes less power, more durable and lighter compared to hard disk drive that sits proudly in your PCs and Laptops today. But not for long. Their time is up. The only downfall of SSD is they are way too expensive. But the research is ON to make them cheaper, and a time will soon come when they will single handedly reign the world until they are overthrown by someone superior. Such is the order of nature. The OLD much make way for the NEW.