All benchmarks measure computation time from JS and Python, respectively. To learn more, check out benchmark methodology.
Size Scaling Summary
| Array Size | Avg Speedup | Best Case | Worst Case | Benchmarks |
|---|---|---|---|---|
| Small (100) | 1.45x | 45.71x | 0.13x | 3513 |
| Medium (1K) | 1.22x | 77.09x | 0.10x | 3513 |
| Large (10K) | 1.42x | 3535.38x | 0.05x | 3502 |
Small (100) — by Category
| Category | Avg Speedup | Count |
|---|---|---|
| creation | 1.63x | 253 |
| arithmetic | 1.10x | 421 |
| math | 1.10x | 292 |
| trig | 1.07x | 216 |
| gradient | 4.68x | 22 |
| linalg | 2.03x | 292 |
| reductions | 2.44x | 521 |
| manipulation | 1.43x | 488 |
| io | 3.52x | 66 |
| indexing | 1.28x | 226 |
| bitwise | 0.68x | 66 |
| sorting | 0.73x | 88 |
| logic | 1.03x | 274 |
| statistics | 3.65x | 28 |
| sets | 1.53x | 121 |
| random | 1.00x | 46 |
| polynomials | 2.36x | 27 |
| fft | 1.25x | 66 |
Medium (1K) — by Category
| Category | Avg Speedup | Count |
|---|---|---|
| creation | 1.35x | 253 |
| arithmetic | 1.30x | 421 |
| math | 1.26x | 292 |
| trig | 1.03x | 216 |
| gradient | 3.78x | 22 |
| linalg | 1.99x | 292 |
| reductions | 1.15x | 521 |
| manipulation | 1.33x | 488 |
| io | 2.14x | 66 |
| indexing | 0.78x | 226 |
| bitwise | 0.61x | 66 |
| sorting | 0.81x | 88 |
| logic | 1.25x | 274 |
| statistics | 1.56x | 28 |
| sets | 0.87x | 121 |
| random | 0.84x | 46 |
| polynomials | 2.30x | 27 |
| fft | 0.72x | 66 |
Large (10K) — by Category
| Category | Avg Speedup | Count |
|---|---|---|
| creation | 3.88x | 253 |
| arithmetic | 1.97x | 421 |
| math | 1.81x | 292 |
| trig | 1.23x | 216 |
| gradient | 5.61x | 22 |
| linalg | 1.67x | 292 |
| reductions | 0.70x | 521 |
| manipulation | 1.47x | 488 |
| io | 2.57x | 55 |
| indexing | 0.59x | 226 |
| bitwise | 1.14x | 66 |
| sorting | 1.01x | 88 |
| logic | 1.99x | 274 |
| statistics | 1.22x | 28 |
| sets | 1.16x | 121 |
| random | 0.88x | 46 |
| polynomials | 2.38x | 27 |
| fft | 1.30x | 66 |