Computer storage has been around for a long time and has taken on many different forms and functions as well as ever-increasing storage amounts. Memory in a computer stores data as a one or zero, often as a state of "on" or "off." This type of storage can be achieved in many ways: in its most basic form, expressed as dots on a piece of paper, for example. Likewise, data can be stored in an electromechanical way using on-or-off light switches.
Electronic tubes were the first type of electrical memory, and later, magnetic core memory that utilized a grid of electrical wires and ferrite rings at the intersections. Since those times computer memory has progressed quite a lot. Today we have many ways of storing an ever-expanding amount of data.
Two Distinctions of Memory
There are two basic distinctions of memory. One is volatile memory where the data is lost as soon as power is removed, and one is non-volatile memory that can store the data without power.
Magnetic core memory was one of the earliest forms of electronic memory and was comprised of a grid of wires with ferrite rings at the intersections. Magnetic core memory is physically very large but allowed for fast random access to the contents.
Random Access Memory
The more modern random access memory (RAM) utilizes silicon transistors and capacitors to store data but is 
done so in a grid-like configuration with a transistor-capacitor pair at each intersection of the grid. The advantage of the grid architecture is that it allows any specific bit of information to be read at any time. Any point on the grid can be read by interrogating the two lines which will read the spot at which they intersect. This type of volatile memory is very fast but the downfall is the capacitors lose their charge over time so the data must constantly be refreshed. This aspect of the memory is called dynamic RAM or DRAM.
By contrast, static random access memory (SRAM) does not need to be refreshed. Because of this, SRAM is faster because it doesn’t require the time necessary to refresh each bit. SRAM is also more expensive and not used as often. Different types of SRAM may lose the memory after power is removed or they may not.
Non-volatile RAM memory or NVRAM is a class of memory that also has a grid architecture, but the data is retained even after the power is removed. Flash memory is an example of a type of NVRAM that utilizes a special type of metal oxide semiconductor field effect transistor (MOSFET) to store data. Other types of NVRAM are ferroelectric RAM (FeRAM) and magnetoresistive ram (MRAM).
Read-only Memory
ROM memory is different from RAM both in operation and use. ROM is non-volatile so it does not lose what is stored without power. While the name suggests that it is read-only and certain types are, more often it is read-only during normal operation but can be written under the special circumstances. ROM also covers a wide array of memory types.
Programmable read-only memory (PROM) is a type of ROM that is typically programmed once and can’t be changed after that. This type of memory is used in things like firmware and RFID chips. It is often used in hardware that has a dedicated purpose that will not change.
EPROM memory is similar to PROM with one key difference. It can be erased and reprogrammed although 
this is not expected to happen often. The memory chips have an optical window on them which must be exposed to UV light which will erase the memory and allow them to be reprogrammed. With this method, the memory must be completely erased before any new information is written.
Since the UV light method is somewhat inconvenient, electrically erasable programmable read-only memory (EEPROM) was developed. Erasing and programming are typically done with a voltage higher than normal operation.
Flash memory like the kind used in SD cards and thumb drives common today was the next step, derived from similar technology that EEPROM uses. Unlike an EPROM or EEPROM flash memory doesn’t require the entire drive to be erased before new data can be stored. The memory uses two kinds of logic gates (Nand and Nor) as its basis depending on the operation. One kind of logic gate allows blocks of memory to be selectively erased and written while the other allows individual bytes. Flash memory has the disadvantage of having a limited number of erase and write cycles before it fails, typically without warning.
Hard Drives and Other Methods
Hard drives are another form of memory often used in computers that come in both electromagnetic and solid state (SSD) versions. Electromagnetic hard drives use spinning disks that have been coated with magnetic material. With the use of magnetic heads that float over the disks, information is read and written to the magnetic material. SSD’s use flash memory in place of the magnetic disks, which makes them faster but also more expensive per the amount of storage.
Tap
Source: Robert Jacek Tomczak CC BY-SA 3.0 e drives are typically used for backup although less so as other forms of storage have become available. As the name suggests a tape that has been coated with a magnetic material is used so that when the tape passes by a tape head, data can be read and written. Tape drives are slow but can be an inexpensive method of storage.
Lastly are optical drives such as CDROM and DVD drives. These use an optical- based disk, which can be written and read with a laser. These are typically used as write-once, but some are designed to be erased and rewritten. In writing, a laser burns pits in the optical disk that do not reflect light. The pits and reflective regions represent ones and zeros. The same method is used for both CDs and DVDs, the difference being that DVDs use thinner tracks to fit more information.
While there are other methods of data storage, these are the major ones. As memory gets larger and cheaper, it will be interesting to see what the next type of memory will be.
