alp_rd: speedup #7064
+96
−51
Merged
alp_rd: speedup #7064
CodSpeed HQ / CodSpeed Performance Analysis
succeeded
Apr 1, 2026
Performance Gate Passed
⚡ 1 improved benchmark
✅ 1097 untouched benchmarks
🆕 24 new benchmarks
⏩ 1530 skipped benchmarks1
Performance Changes
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ⚡ | Simulation | alp_rd_decompress_f64 |
2.3 ms | 1.2 ms | +92.57% |
| 🆕 | Simulation | compress_rd[f32, (100000, 0.01)] |
N/A | 3.6 ms | N/A |
| 🆕 | Simulation | compress_rd[f64, (100000, 0.01)] |
N/A | 4.7 ms | N/A |
| 🆕 | Simulation | compress_rd[f64, (100000, 0.1)] |
N/A | 4.7 ms | N/A |
| 🆕 | Simulation | compress_rd[f64, (10000, 0.01)] |
N/A | 491.5 µs | N/A |
| 🆕 | Simulation | compress_rd[f64, (10000, 0.1)] |
N/A | 493 µs | N/A |
| 🆕 | Simulation | compress_rd[f32, (100000, 0.1)] |
N/A | 3.6 ms | N/A |
| 🆕 | Simulation | compress_rd[f32, (100000, 0.0)] |
N/A | 3.6 ms | N/A |
| 🆕 | Simulation | compress_rd[f32, (10000, 0.01)] |
N/A | 381.4 µs | N/A |
| 🆕 | Simulation | compress_rd[f64, (100000, 0.0)] |
N/A | 4.7 ms | N/A |
| 🆕 | Simulation | compress_rd[f64, (10000, 0.0)] |
N/A | 491.5 µs | N/A |
| 🆕 | Simulation | compress_rd[f32, (10000, 0.0)] |
N/A | 382.1 µs | N/A |
| 🆕 | Simulation | compress_rd[f32, (10000, 0.1)] |
N/A | 382.7 µs | N/A |
| 🆕 | Simulation | decompress_rd[f32, (100000, 0.01)] |
N/A | 757.4 µs | N/A |
| 🆕 | Simulation | decompress_rd[f32, (100000, 0.0)] |
N/A | 757.7 µs | N/A |
| 🆕 | Simulation | decompress_rd[f32, (10000, 0.0)] |
N/A | 145.4 µs | N/A |
| 🆕 | Simulation | decompress_rd[f32, (10000, 0.01)] |
N/A | 106.1 µs | N/A |
| 🆕 | Simulation | decompress_rd[f64, (10000, 0.01)] |
N/A | 156.9 µs | N/A |
| 🆕 | Simulation | decompress_rd[f64, (10000, 0.0)] |
N/A | 156.7 µs | N/A |
| 🆕 | Simulation | decompress_rd[f64, (10000, 0.1)] |
N/A | 156.5 µs | N/A |
| ... | ... | ... | ... | ... | ... |
ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.
Comparing ji/alp-rd-speedup (7847609) with develop (db00b43)
Footnotes
-
1530 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩
Loading