Storage Devices: Primary vs Secondary Memory, Examples and Units

A clear guide to what storage devices are, how primary and secondary memory differ, examples, how storage evolved, and how data is measured.

By Aaron JoelPublished 7 min read1 source listed

A storage device keeps your data and programs so a computer can use them later. This guide explains what storage devices are, how primary memory differs from secondary storage, common examples, how storage has changed over time, and how data size is measured.

Definition

A storage device is hardware that holds digital data and programs, either temporarily or permanently, so a computer can read or write them when needed.

  • Why it is needed: the CPU can only work on data it can reach quickly, and computers must keep programs, documents and the operating system even when the power is off.
  • Basic operations: write (save data), read (retrieve data), erase/overwrite (replace data).
  • Data is stored as bits: every file, whether text, image, video or program, is a pattern of 0s and 1s.
  • Key properties used to compare devices:
    • Capacity: how much data it holds
    • Speed: how fast data is read and written
    • Volatility: whether data survives power loss
    • Portability: whether it can be carried and moved between computers
    • Cost per GB and durability

Terms to keep apart: storage medium is the material that holds the data (a disk platter, a flash chip). Storage device is the drive that reads and writes that medium (a hard disk drive, a card reader).

Primary and Secondary

Storage is split by how close it sits to the CPU: primary memory is read directly by the CPU while programs run, and secondary storage keeps data long term. (Both pairs of names, “memory” and “storage”, are used in exams; follow the wording of the question.)

Feature Primary memory (main memory) Secondary storage (auxiliary storage)
CPU access Direct, through the memory bus Indirect, via I/O controllers; data is loaded into RAM first
Speed Very fast (nanoseconds) Slower (microseconds to milliseconds)
Volatility RAM is volatile; ROM is non-volatile Non-volatile
Capacity Small (GBs) Large (hundreds of GB to TBs)
Cost per GB High Low
Purpose Holds running programs and the data in use Holds the OS, programs and files permanently
Examples RAM, ROM, cache, registers HDD, SSD, USB flash drive, SD card, CD/DVD, magnetic tape

Tertiary storage (sometimes added as a third level) means offline archive media such as tape libraries and optical jukeboxes, used for backups and rarely accessed data.

Order from fastest to slowest: registers, cache, RAM, SSD, HDD, optical disc, tape. Going down the list, speed falls while capacity rises and cost per GB falls.

Examples

Primary memory

Device Type What it does
Registers Volatile Tiny storage inside the CPU; holds the instruction and data being processed right now
Cache (L1, L2, L3) Volatile Very fast memory on or near the CPU; keeps frequently used data to avoid slower RAM trips
RAM (DRAM, SRAM) Volatile Main working memory; holds running programs and open files
ROM (PROM, EPROM, EEPROM) Non-volatile Holds firmware such as the BIOS/UEFI boot program

Secondary storage

Device Technology Typical use
Hard disk drive (HDD) Magnetic spinning platters Cheap bulk storage in PCs, servers and external drives
Solid state drive (SSD) Flash memory, no moving parts Fast boot drive for laptops and PCs
USB flash drive (pen drive) Flash memory Portable file transfer
Memory card (SD, microSD) Flash memory Cameras, phones, drones
CD, DVD, Blu-ray Optical (laser) Software, movies, backups
Magnetic tape Magnetic tape Long-term archives and enterprise backups
Cloud storage Remote servers over the internet Online backup and file sharing (Google Drive, OneDrive)

Exam tip: classify by technology as well as by name: magnetic (HDD, tape), optical (CD, DVD, Blu-ray) and solid state/flash (SSD, USB, SD card).

Evolution and Limitations

Storage moved from paper cards to magnetic, optical and flash media, with each step giving more capacity, more speed and a smaller size. Dates below are approximate.

Period Technology Typical capacity Main limitations
1890s to 1970s Punch cards and paper tape About 80 characters per card Bulky, slow, one-way reading, easily damaged, no rewriting
1950s Magnetic tape (UNIVAC I, 1951) Few MB per reel Sequential access only, so finding data means winding the tape
1956 onward Hard disk drive (IBM 350 RAMAC) About 3.75 MB at launch, now many TB Moving parts, so slower, noisy, fragile and prone to mechanical failure
1970s to 2000s Floppy disk (8-inch, 5.25-inch, 3.5-inch) 80 KB up to 1.44 MB Tiny capacity, slow, easily corrupted by magnets and dust
1980s to 2000s Optical discs (CD about 700 MB, DVD 4.7 GB, Blu-ray 25 GB per layer) MB to tens of GB Scratches damage data, slow writing, mostly write-once or limited rewrites
Late 1990s onward Flash storage (USB drive, SD card, SSD) GBs to several TB Costs more per GB than HDD, limited write cycles per cell
2000s onward Cloud storage Effectively unlimited, paid by usage Needs internet, ongoing subscription cost, privacy and security concerns

Limitations common to all storage

  • Capacity versus cost: more space always costs more, and fast storage costs the most per GB.
  • Speed gap: even the fastest SSD is far slower than RAM, which is why caches and RAM are used between the CPU and storage.
  • Durability: magnetic media lose data over time and near magnets, optical discs scratch, and flash cells wear out after many writes.
  • Failure and data loss: every device can fail, so regular backups are needed.
  • Security: portable media (pen drives) are easy to lose or steal, and data should be encrypted.

Trend: devices have become smaller, faster, cheaper per GB and more portable, while capacity keeps rising. HDDs remain popular for bulk storage, and SSDs have taken over as the main drive in most new laptops.

Measurements

The smallest unit of data is the bit (binary digit, 0 or 1). Larger units are built from bits, and in computing each step up is usually a factor of 1024 (2^10).

Unit Symbol Equals Example
Bit b 0 or 1 One on/off signal
Nibble 4 bits One hexadecimal digit
Byte B 8 bits One character, such as “A”
Kilobyte KB 1024 bytes A short text file
Megabyte MB 1024 KB A photo or an MP3 song
Gigabyte GB 1024 MB A movie or a USB drive
Terabyte TB 1024 GB A large hard disk
Petabyte PB 1024 TB A data centre
Exabyte EB 1024 PB Global-scale storage

Decimal and binary prefixes

Manufacturers use powers of 1000 (1 GB = 1,000,000,000 bytes), while operating systems often count in powers of 1024. This is why a drive sold as 500 GB shows as about 465 GB in Windows. The standard binary names are KiB, MiB, GiB and TiB, where 1 GiB = 1024 MiB. Most school syllabuses use 1 KB = 1024 bytes unless told otherwise.

Conversion rules and worked examples

  • Larger to smaller unit: multiply by 1024 (3 MB = 3 × 1024 = 3072 KB).
  • Smaller to larger unit: divide by 1024 (2048 KB = 2048 ÷ 1024 = 2 MB).
  • Bytes to bits: multiply by 8 (1 KB = 1024 × 8 = 8192 bits).
  • Capacity question: a 16 GB card holds 16 × 1024 = 16,384 MB, so it fits about 3,276 photos of 5 MB each.
  • Speed versus size: internet speed is in bits per second (Mbps) and file size is in bytes, so divide Mbps by 8 for MB/s (an 80 Mbps line downloads about 10 MB/s).

Memory trick: B, KB, MB, GB, TB, PB, EB: “Be Kind, My Great Teacher, Please Explain”.

References

  1. Computer Memory - GeeksforGeeks