This guide provides a concise overview of storage and compute estimation, common sizes, and solutions by scale for system design interviews and real-world applications.

---

1. Storage Estimation

Key Estimations for Storage:

Item	Avg Size	Examples
Text Message	50 bytes	Chat messages
Tweet	200 bytes	Includes hashtags, metadata
User Profile	1-5 KB	Profile info, settings, and preferences
Image (compressed)	100-500 KB	JPG/PNG, low to medium resolution
Request Payload	50 KB	POST/PUT Request
Video (1 min)	5-10 MB	Compressed, e.g., 720p
Log Entry	100 bytes	Timestamp, log level, message
Sensor Data Point	100 bytes	IoT data with metadata
Machine Learning Model	100 MB - 1 GB	Simple to complex models

Storage Solutions by Scale:

Data Volume	Suggested Solution	Examples
< 10 GB	Single DB Instance	MySQL, PostgreSQL
10 GB - 1 TB	Partitioned DB or file storage	PostgreSQL, AWS S3
1 TB - 100 TB	Distributed storage or sharded DB	MongoDB, Cassandra, HDFS
> 100 TB	Data lake or distributed file system	AWS S3, Hadoop HDFS

---

2. Compute Estimation

Key Estimations for Compute:

• QPS Calculation: \( \text{QPS} = \frac{\text{DAU} \times \text{Actions/User}}{86,400} \)
• Peak Load: Generally 3-5x the average QPS for consumer applications.
• Example: 1M DAU with 20 actions/day → \( \approx 20 \) QPS avg, ~100 QPS peak.

Compute Solutions by Scale:

QPS	Suggested Solution	Examples
< 100 QPS	Single server or small cluster	EC2 instance, Kubernetes pod
100 - 1000 QPS	Autoscaling cluster	AWS ECS, GCP Compute Engine
1000 - 10,000 QPS	Load-balanced cluster with caching	Redis, Kubernetes, Load balancer
> 10,000 QPS	Distributed, multi-region setup	Global load balancers, sharded DB

---

Common Use Case Scenarios

Scenario	Estimated Usage	Solution
1M DAU Chat App	20 QPS peak, 1 GB/day	Autoscaling, Redis cache
50M Tweets/Day	600 QPS avg, 2.5 TB/year	Sharded DB, multi-region storage
10K QPS E-commerce	500 MB/s data throughput	Load balancing, distributed caching
1 PB Data Lake	100 TB/month growth	Data lake, e.g., S3, Hadoop HDFS
100M Video Views/Day	1 PB storage monthly, 2 Gbps peak	CDN, global load balancer
5M User Transactions/Day	50 QPS avg, 5 TB storage/year	Sharded RDBMS, caching, high consistency

---

Quick Calculation Tips

• Time Conversions:
• 1 day ≈ \(10^5\) seconds, 1 month ≈ \(10^6\) seconds
• Bytes Conversions:
• 1 KB ≈ \(10^3\) bytes, 1 MB ≈ \(10^6\) bytes, 1 GB ≈ \(10^9\) bytes, 1 TB ≈ \(10^{12}\) bytes

Bytes to Powers of 10 Conversion Table

Unit	In Powers of 10	Approximation
Kilobyte (KB)	\( 10^3 \)	1 thousand bytes
Megabyte (MB)	\( 10^6 \)	1 million bytes
Gigabyte (GB)	\( 10^9 \)	1 billion bytes
Terabyte (TB)	\( 10^{12} \)	1 trillion bytes
Petabyte (PB)	\( 10^{15} \)	1 quadrillion bytes
Exabyte (EB)	\( 10^{18} \)	1 quintillion bytes