7.1Hierarchy
The idea
The memory hierarchy stacks memories from tiny-and-fast to huge-and-slow so the CPU usually finds data in the fast ones.
Think of it like this: Pocket, desk, bookshelf, library: ever bigger, ever slower.
Why it exists
No technology is fast, large and cheap at the same time; a hierarchy gives the illusion of all three.
How it works
- Registers → L1 → L2/L3 → DRAM → disk.
- Going down: capacity ↑, access time ↑, cost per bit ↓.
- It works thanks to locality of reference.
- Data moves between levels in blocks or pages.
Worked example: Typical times
Register ~0.3 ns, L1 ~1 ns, DRAM ~60–100 ns, SSD ~100 µs, HDD ~5–10 ms.
Now play with it
- →Click each level to see its typical size and speed.
Going down: capacity ↑, access time ↑, cost per bit ↓, frequency of access ↓.
Rough typical figures, to build intuition (ratios matter, not exact values).
Check yourself
Going down the memory hierarchy, which quantity decreases?
Exam notes (the dense version)
Registers → L1 → L2/L3 → Main memory (DRAM) → Disk. Going down: bigger, slower, cheaper per bit. Works because of locality.
Previous year questions: Memory Hierarchy & Main Memory (16)
PYQs are grouped by topic, so this is the full set for Memory Hierarchy & Main Memory; try the ones that match this subtopic. Open all 16 on the topic page.