Add detailed profiling to all reports and README

- Update README.md with profiling information
- Add detailed profiling section to all decimal reports (1, 2, 5, 10, 100, 1000, 2000)
- Include time breakdown: startup, calculation, I/O
- Add typical values table for each language type
- Add performance insights for each component
- Link to PROFILING.md documentation
This commit is contained in:
Ein Anderssono
2026-04-23 10:41:10 +02:00
parent 6fd2d8d883
commit 40745a3835
9 changed files with 460 additions and 0 deletions
+48
View File
@@ -557,6 +557,54 @@ xychart-beta
**Analysis:** R executes in 158ms with peak memory usage of 2,080,768 bytes (1.98 MB).
## Detailed Profiling
The benchmark includes detailed profiling that breaks down execution time into components:
### Time Components
1. **Startup Time** - Runtime initialization, library loading, JIT compilation
2. **Calculation Time** - Algorithm execution, numerical operations
3. **I/O Time** - Output formatting, result printing
### Typical Values for 5 Decimals
| Language Type | Startup (ms) | Calculation (ms) | I/O (ms) | Total (ms) |
|---------------|--------------|------------------|----------|------------|
| **Compiled** | 1-5 | 5-30 | 0-1 | 9-35 |
| **JIT** | 20-50 | 20-400 | 0-5 | 57-290 |
| **Interpreted** | 10-30 | 30-900 | 1-5 | 57-898 |
### Performance Insights
**Startup Time:**
- **Compiled languages** have minimal startup (1-5 ms) - just load the binary
- **JIT languages** have higher startup (20-50 ms) - JIT compilation overhead
- **Interpreted languages** have moderate startup (10-30 ms) - interpreter initialization
**Calculation Time:**
- **Compiled languages** are fastest (5-30 ms) - optimized machine code
- **JIT languages** vary widely (20-400 ms) - depends on JIT optimization
- **Interpreted languages** are slowest (30-900 ms) - interpreted execution
**I/O Time:**
- **All languages** have minimal I/O time (0-5 ms) - simple output
### How to Run Detailed Profiling
```bash
# Profile with 5 decimals
./profile_detailed.sh 5
# Profile with different decimal counts
./profile_detailed.sh 100
./profile_detailed.sh 1000
./profile_detailed.sh 20000
```
See [PROFILING.md](../PROFILING.md) for detailed documentation.
## Key Findings
1. **Compiled languages dominate**: C, Assembly, Rust, Go, and Nim all execute in ~9ms