Suggest an edit

Improve this article

Refine the answer for “Why are I/O operations considered slow?”. Your changes go to moderation before they’re published.

Approval required

Content

What you’re changing

Title (EN)

Short answer (EN)

Shown above the full answer for quick recall.

Answer (EN)

**I/O operations are considered "slow"** because they **depend on external devices or networks**, not on the processor.
"Slow" here means slow **compared to operations in RAM or the CPU**.

## 1. What happens during I/O

When a program, say, reads a file:

1. The **processor** hands the operating system a request: "read data from disk".
2. The OS passes the task to a **device driver**.
3. The **physical disk** (SSD or HDD) has to:
  - locate the right sector,
  - read the data,
  - pass it back through the system bus into memory.
4. Only then does the program get the result.

This can take **milliseconds**, which for the CPU is **an eternity** (since the processor operates in nanoseconds).

## 2. Typical examples of "slow" I/O

| I/O type | Average latency | Why it's slow |
|---|---|---|
| **Disk (HDD/SSD)** | milliseconds | physically reading data off the media |
| **Network (HTTP, TCP)** | milliseconds → seconds | waiting for network transfer, a server response |
| **Database** | milliseconds | network delays + running the query on the database side |
| **Filesystem** | milliseconds | talking to the OS, accessing file descriptors |

## 3. Comparison with CPU and memory

| Operation type | Time (roughly) |
|---|---|
| CPU (arithmetic, memory) | nanoseconds (10⁻⁹ s) |
| RAM operations | tens of nanoseconds |
| An SSD | hundreds of microseconds → milliseconds |
| An HTTP request | milliseconds → hundreds of milliseconds |

That's a difference of **millions of times** between computing in memory and reading from disk/network.

## 4. Why this matters in Node.js

Node.js is **single-threaded**.
If I/O operations ran **synchronously (blocking)**, then:

- the thread would be **stuck waiting** on the disk or network;
- other requests would **hang** until the operation finished.

That's why Node.js uses **non-blocking asynchronous I/O**, operations run in the background, and once ready, they trigger a **callback** (through the Event Loop).

## Summary

> **I/O operations are considered slow** because they require interacting with external devices or networks, where delays (milliseconds) are orders of magnitude larger than in-memory operations (nanoseconds).
> Node.js solves this with **asynchronous, non-blocking I/O**, so it never sits idle while waiting.

Markdown · drag & drop images · ⌘B / ⌘I shortcuts370 words

For the reviewer

Note to the moderator (optional)

Visible only to the moderator. Helps review go faster.