Internal Structure of a Hash Table

Watch the video explanation ➔

Hash Tables are implemented through simple arrays, but how?

Hash Tables are so powerful, that OOP-based languages internally use them to power Classes and site members. Symbol tables that hold the variables mapped to a memory location are also powered through hash tables.

They are designed to provide constant-time key-based insertion, update, and lookups while being space efficient at all times.

Core Ideas to construct Hash Tables

  • convert application keys to wide-ranged (INT32) hash keys
  • convert hash keys to a smaller range

Application Keys to Hash Keys

Hash Tables should support storing an object as a key and to power that the keys are first hashed to a big integer range (provided by the user) typically INT32. This hash key is then further used to decide how and where the KV pair would be stored in the data structure.

Naive Implementation

A naive implementation of a Hash Table would be to create an array of length INT32. To store the KV in it, we pass the key through the hash function, spitting out an integer. We use this key and store the KV pair at this index in the array.

Although this would give us constant time insertion, update, and lookups, it is highly space in-efficient, as we would need to allocate at least 4 * INT32 = 16GB of ram to just hold this array, with most of the slots left empty.

Hash Keys to Smaller Range

This step is designed and introduced to make our Hash Table space efficient. Instead of having a huge array of length INT32, we keep it proportional to the number of keys inserted. For example, if we inserted 4 keys, then our holding array could be around 8 slots big.

To achieve this, we map the hash key into a small range (same as the length of the array) and place our key at that very index. This allows us to remain space-efficient while sporting fast and efficient insertions, updates, and lookups.

Adding more keys

The small limited-size array will not be able to hold numerous keys and hence after a certain stage we would need a larger array to hold the data. This is done by resizing the holding array and is typically made 2x every time it is full enough.

Thus, this two-step implementation allows for near-constant time insertions, updates, and lookups while remaining space efficient.

Here's the video ⤵


Super practical courses, with a no-nonsense approach, are designed to spark engineering curiosity and help you ace your career.

System Design for Beginners

An in-depth, self-paced, and on-demand course that for early engineers to become great at designing scalable, available, and extensible systems at scale.

Details →

System Design Masterclass

A masterclass that helps experienced engineers become great at designing scalable, fault-tolerant, and highly available systems.

Details →

Redis Internals

A course that helps covers Redis internals by reimplementing its core features like - event loop, serialization protocol, pipelining, eviction, and transactions.

Details →

Writings and Videos


Essays and Blogs

Arpit's Newsletter read by 70000+ engineers

Weekly essays on real-world system design, distributed systems, or a deep dive into some super-clever algorithm.