1 of the 102 libraries were not tested in the same version, or not with the same configuration and reference files, in the two runs; see Library Changes below.
The baseline is the newest run of wasm-jit, not the commit the pull request is
based on, so a difference can also come from something merged into wasm-jit since
the pull request was branched.
The two runs
| Branch | Run | Compiler | Machine |
|---|
| Baseline | wasm-jit | 2026-10-06 18:05:05 | v1.28.0-dev.1294+gfd4849f190.rust | ryzen-9950x |
| Pull request | pr/17156 | 2026-10-06 22:34:19 | v1.28.0-dev.1298+g017c65039a.rust | ryzen-9950x |
Summary
| Models compared | 20037 |
| Number of Improvements | 5 |
| Number of Regressions | 1 |
| Number of Performance Improvements | 3 |
| Number of Performance Regressions | 3 |
| Models only in the pull request run | 1 |
| Models only in the baseline run | 0 |
Time
Time spent on the 20037 models both runs have. The last column counts only the 20031 that reached the same phase in both: a model that fails earlier stops paying for the phases it no longer reaches.
| Phase | wasm-jit | pr/17156 | Change | Change, same phase |
|---|
| frontend | 1:17:16 | 1:15:19 | -2.5% | -2.5% |
| backend | 1:48:10 | 1:49:16 | +1.0% | +1.0% |
| simcode | 0:22:32 | 0:22:24 | -0.6% | -0.6% |
| templates | 0:12:13 | 0:11:38 | -4.8% | -4.8% |
| compile | 1:18:47 | 1:15:38 | -4.0% | -4.1% |
| simulate | 7:48:37 | 7:54:54 | +1.3% | +1.1% |
| verify | 0:14:42 | 0:14:32 | -1.1% | -1.1% |
| total | 19:10:48 | 19:07:10 | -0.3% | -0.4% |
Library Changes
| Library | Change |
|---|
| PowerGrids_dev | Version 2.0.0-develop (edaee31c23448708471b0b4124b10fda519c317a) in wasm-jit, 2.0.0-develop (beae7748a855afc478e5abcfcb8625358ec175ac) in pr/17156 |
Models Affected
Models Only in One of the Runs
| Library | Model | Where |
|---|
| PowerGrids_dev | PowerGrids.Examples.Tutorial.IslandOperation.PowerFlowLocalInitialization | only in pr/17156 |