- Published on
System Design: Must-Know Numbers
- Authors
- Name
- Loi Tran
Introduction
System Design by the Numbers: A Mental Model for Latency, Load, and Storage follows.
✅ Basic Computer Science Constants
| Concept | Value |
|---|---|
| 1 Byte | 8 bits |
| 1 KB | 1024 Bytes |
| 1 MB | 1024 KB (~1 million bytes) |
| 1 GB | 1024 MB (~1 billion bytes) |
| 1 TB | 1024 GB |
✅ Time Units
| Unit | Symbol | Meaning |
|---|---|---|
| ns | nanosecond | 1 billionth of a second (10⁻⁹ s) |
| µs | microsecond | 1 millionth of a second (10⁻⁶ s) |
| ms | millisecond | 1 thousandth of a second (10⁻³ s) |
| s | second | 1 second (10⁰ s) |
✅ Latency Benchmarks (Rule of Thumb)
These help you reason about how long certain operations take.
| Operation | Approx Time |
|---|---|
| L1 cache reference | ~0.5 ns |
| L2 cache reference | ~7 ns |
| Main memory (RAM) access | ~100 ns |
| SSD read (local) | ~100 µs |
| HDD read (seek+read) | ~10 ms |
| Roundtrip within same datacenter | ~0.5 ms |
| Roundtrip between datacenters | ~100 ms |
✅ Bandwidth Estimates
| Operation | Speed (ballpark) |
|---|---|
| 1 Gbps network | ~125 MB/s |
| 10 Gbps network | ~1.25 GB/s |
| SATA SSD read/write | ~500 MB/s |
| NVMe SSD read/write | ~3–5 GB/s |
| RAM bandwidth | ~10–30 GB/s |
✅ Common Design Numbers
| Context | Estimate |
|---|---|
| Disk seek time | 5–10 ms |
| DB write latency (local) | ~1–10 ms |
| Kafka throughput per broker | ~100 MB/s (varies) |
| Redis GET latency | ~0.5–1 ms |
| Web request latency target | < 100 ms (good), <1s (acceptable) |
✅ User Behavior Numbers
| Context | Estimate |
|---|---|
| Tweets per second | ~6,000 TPS (2023 avg) |
| Facebook photo uploads | ~300 million/day (~3.5k/s) |
| 1 million users: 1% concurrency | ~10,000 concurrent users |
✅ Storage Size Approximations
| Item | Size (approximate) |
|---|---|
| Text message | ~1 KB |
| Tweet | ~1–2 KB (with metadata) |
| Web page HTML | ~50–200 KB |
| Image (compressed JPG) | ~100 KB – 2 MB |
| Video (HD 1080p, 1 min) | ~60–100 MB |