Revert "fix: build CUDA kernels as multi-arch fatbin with PTX fallback" #8055
+36
−40
CodSpeed HQ / CodSpeed Performance Analysis
failed
May 21, 2026 in 0s
Performance Regression: -1.68%
⚠️ Unknown Walltime execution environment detected
Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.
For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.
⚡ 2 improved benchmarks
❌ 1 regressed benchmark
✅ 1234 untouched benchmarks
Warning
Please fix the performance issues or acknowledge them on CodSpeed.
Performance Changes
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ⚡ | Simulation | baseline_eq[16, 65536] |
287.6 µs | 259.6 µs | +10.76% |
| ⚡ | Simulation | baseline_lt[16, 65536] |
302.7 µs | 274.7 µs | +10.17% |
| ❌ | Simulation | fast_lt_out_of_range[4, 65536] |
204.3 µs | 262.3 µs | -22.11% |
Tip
Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.
Comparing revert-8047-fix-cuda-ptx-gpu-invalidation (68dc253) with develop (c54ce7e)
Loading