-
Notifications
You must be signed in to change notification settings - Fork 12
Benchmarks
Ashton Meuser edited this page Sep 1, 2023
·
2 revisions
Comparison of GDScript, GDNative, and Wasm (n=1000, p95). The following benchmarks were run on macOS 12.6.3, 16GB RAM, 2.8 GHz i7. The project was exported to avoid GDScript slowdown likely caused by performance monitoring. Speedup figures for both GDNative and Wasm are relative to GDScript. The benchmarks used are 1) a DP Nth Fibonacci number and 2) a Sieve of Atkin.
| Benchmark | GDScript | GDNative | GDNative Speedup WRT GDScript |
Wasm | Wasm Speedup WRT GDScript |
|---|---|---|---|---|---|
| Fibonacci (index=20) | 6.0 µs | 1.0 µs | 6.0x | 4.0 µs | 1.5x |
| Fibonacci (index=50) | 13.0 µs | 1.0 µs | 13.0x | 4.0 µs | 3.3x |
| Sieve (limit=1000) | 704.1 µs | 4.0 µs | 176.0x | 30.0 µs | 23.5x |
| Sieve (limit=10000) | 6331.8 µs | 52.0 µs | 121.8x | 134.1 µs | 47.2x |
| Sieve (limit=100000) | 62454.1 µs | 341.0 µs | 183.1x | 1053.1 µs | 59.3x |