Architecture Cost Decision Tool
Compare AWS database architectures for your scale
Inputs
Derived Metrics
Events per Day
24,000,000
Events per Month
720,000,000
Events per User Lifecycle
4,380
Storage per User (MB)
4,380
Users Retained (4 yrs)
24,000,000
Total Storage (GB)
102,656,250
Average RPS
277.78
Peak RPS
1,333.33
Required vCPUs
19.05
Rounded vCPUs
20
Scenario Comparison
RDS Only Architecture
| Component | Monthly Cost |
|---|---|
| RDS Storage | $11,805,468.75 |
| RDS Instance | $3,500.00 |
| IOPS & Backups | $1,000.00 |
| Compute (ECS) | $591.01 |
| Cache | $120.00 |
| Queue | $30.00 |
| Security | $200.00 |
| Monitoring | $250.00 |
| Networking | $300.00 |
| TOTAL | $11,811,459.76 |
When to Use
Best for applications requiring strong ACID guarantees, complex queries with JOINs, and traditional relational data models. Ideal when your team has deep SQL expertise.
Strengths
- Mature ecosystem with extensive tooling
- Complex query support (JOINs, aggregations, window functions)
- Strong consistency and ACID transactions
- Familiar to most development teams
Trade-offs
- Storage costs scale linearly with data volume
- Vertical scaling limits (instance size constraints)
- Higher operational overhead for very large datasets
- Less cost-effective at massive scale compared to NoSQL
DynamoDB Only Architecture
| Component | Monthly Cost |
|---|---|
| DynamoDB Storage | $25,664,062.50 |
| DynamoDB PITR | $20,531,250.00 |
| Read/Write Requests | $200.00 |
| Compute | $300.00 |
| Cache | $120.00 |
| Queue | $30.00 |
| Security | $200.00 |
| Monitoring | $250.00 |
| Networking | $300.00 |
| TOTAL | $46,196,712.50 |
When to Use
Best for high-scale applications with simple access patterns, key-value lookups, and predictable query patterns. Ideal for serverless architectures and global distribution.
Strengths
- Virtually unlimited horizontal scaling
- Single-digit millisecond latency at any scale
- Fully managed with zero operational overhead
- Pay-per-request pricing option for variable workloads
- Global tables for multi-region replication
Trade-offs
- Storage costs significantly higher than RDS ($0.25/GB vs $0.115/GB)
- PITR backup adds 80% overhead to storage costs
- Limited query flexibility (no JOINs, complex aggregations)
- Requires careful data modeling and access pattern planning
- Total cost can be 4x higher than RDS for this workload
Hybrid Architecture (Hot/Warm/Cold)
| Component | Monthly Cost |
|---|---|
| Hot+Warm RDS Storage | $14,765,625.00 |
| Cold S3 Storage | $87,890,625.00 |
| RDS Instance | $1,698,046.88 |
| IOPS & Backups | $2,109,375.00 |
| Compute (ECS) | $2,500.00 |
| Cache | $600.00 |
| Athena / ETL | $591.01 |
| Queue | $120.00 |
| Security | $150.00 |
| Monitoring | $30.00 |
| Networking | $200.00 |
| TOTAL | $3,811,612.88 |
When to Use
Best for applications with time-based access patterns where recent data needs fast access but historical data is rarely queried. Ideal for analytics, compliance, and audit use cases.
Strengths
- Optimized cost-to-performance ratio
- Hot data in RDS for complex queries
- Cold data in S3 reduces storage costs by 80%
- Athena enables SQL queries on S3 without ETL
- Flexible data tiering based on access patterns
- Can achieve 93% cost reduction vs DynamoDB
Trade-offs
- Increased architectural complexity
- Requires automated data lifecycle management
- Cold data queries have higher latency
- ETL pipeline adds operational overhead
- Need clear data tiering policies
- Cross-service integration and monitoring complexity