Storage Hierarchy Comparison

About this interactive

What you’re seeing: the eight places a computer can keep a byte, drawn as a pyramid — CPU registers at the narrow top, then L1, L2 and L3 cache, RAM, SSD, HDD, and optical or tape at the wide base. Hover any tier for its access time, capacity, cost per gigabyte, whether it survives a power cut, and what it is made of. Why it matters: the hierarchy is normally memorised as an order, and the order is the least interesting part of it — the gaps are what explain why a cache miss hurts, why more RAM fixes a slow machine, and why nothing runs from tape. Top to bottom is about eleven orders of magnitude, and the steps are nowhere near equal. How to use it: click a tier to park a byte there, then drag the journey slider (or press Play the climb) and watch it copied up through every tier above, paying each one’s latency on the way. Read the running total at the end — climb from the HDD and about 99% of the trip was the disk alone. Every figure is a round approximation on purpose; the size of the steps is the fact worth carrying, not the digits.

How to use this tool

A pyramid of the eight places a computer can keep a byte, fastest and smallest at the top, slowest and largest at the bottom: CPU registers, L1, L2 and L3 cache, RAM, SSD, HDD, and optical disc or tape.

Hover a tier (or tab to it and press Enter) and the panel beside the pyramid fills in with that tier’s access time, how much of it a machine typically has, what a gigabyte of it costs, whether it keeps your data when the power goes, and what it is physically made of.

Click a tier to park a byte there. Then drag the Journey slider — or press Play the climb — and watch that byte copied up through every tier above it, paying each one’s latency on the way, because a tier can only read from the tier directly below it. The callouts fire as it passes each tier and the running total builds underneath.

Read the IF 1 NS = 1 S figure as you go: it is the same wait with one nanosecond stretched to one second, which is the only way these numbers become imaginable. And watch what happens to the total at the end of a climb from the HDD or the tape — almost all of it is the first hop. That is the whole reason caches exist.

Every figure here is an order-of-magnitude approximation, written with a “~” or a range on it. Real parts vary; the size of the steps between tiers does not.

What you'll learn

Order the tiers of the storage hierarchy from CPU registers to archival tape and explain what each step down trades away in speed to gain in capacity and cost per gigabyte
Estimate the access time of each tier to the right order of magnitude, and explain why the slowest tier a request touches accounts for essentially all of its total latency

Aligned to

CompTIA Tech+

2.5 Compare and contrast storage types.

2.3 Explain the purpose of common internal computing components.

1.3 Illustrate the basics of computing and processing.

Key terms

Storage: The component of a computer that retains data for later use, such as a USB drive, solid state drive, or SD card.
Cache: A small, high-speed storage area that holds frequently or recently accessed data so it can be retrieved faster than fetching it from a slower source; CPU caches (L1, L2, L3) reduce processor wait time for main memory, while browser caches store local copies of web resources to speed up page loads.
CPU Cache: A small, fast memory built into the CPU chip that temporarily holds data immediately before and after processing.
Static Random-access Memory: A type of volatile RAM that uses bistable flip-flop circuits to store data without requiring periodic refresh, making it faster and more expensive than DRAM; commonly used for CPU cache memory (L1, L2, L3).
Random Access Memory: Volatile memory that temporarily stores data and program instructions the CPU is actively using, allowing fast read and write access from any memory location. RAM contents are lost when power is removed, unlike persistent storage.
Dynamic Random-Access Memory: A type of volatile RAM that stores each bit in a separate capacitor cell, requiring periodic electrical refreshing to retain data; it is the most common type of main system memory in computers.
Volatile Memory: A category of storage that requires continuous electrical power to retain its contents; when power is removed, all data is lost. RAM is the primary example, used to hold the programs and data a CPU is actively working with.
Non-Volatile Memory: A category of storage that retains data even when power is removed, used for persistent storage of the operating system, applications, and user data; SSDs and ROM are examples.
Solid State Drive: A storage device that uses flash memory chips with no moving mechanical parts to store data persistently, offering faster read/write speeds, lower latency, and greater durability than traditional hard disk drives. SSDs connect via SATA, NVMe, or M.2 interfaces.
Hard Disk Drive: A traditional storage device that uses rotating magnetic platters and read/write heads to store and retrieve data persistently. HDDs offer high-capacity storage at lower cost per gigabyte than SSDs, but with slower speeds and moving mechanical parts.
Optical Storage: A storage technology that uses laser light to read and write data on reflective disc surfaces, such as CDs and DVDs.
Magnetic Tape: A sequential magnetic storage medium consisting of a thin strip of magnetizable material used to store data as a series of polarized bits.

Topics

Storage Storage Hierarchy Cache Memory Ram Ssd Hdd Latency Cost Per Gigabyte Volatility

About TechKnowSurge

TechKnowSurge builds IT and cybersecurity professionals through hands-on, concept-first training built around real understanding — not memorization. Free interactive tools, structured programs, and 25+ years of real-world experience, all in one place.

Explore free tools and programs →