Change the default routing_method to sabre (#12806) (#12847)

* change the default routing_method to sabre

* reno

* Update releasenotes/notes/sabre_level0-1524f01965257f3f.yaml

Co-authored-by: Matthew Treinish <[email protected]>

---------

Co-authored-by: Matthew Treinish <[email protected]>
(cherry picked from commit b985b96)

Co-authored-by: Luciano Bello <[email protected]>

Author: mergify[bot]

qiskit/transpiler/preset_passmanagers/level0.py

@@ -48,7 +48,7 @@ def level_0_pass_manager(pass_manager_config: PassManagerConfig) -> StagedPassMa
     initial_layout = pass_manager_config.initial_layout
     init_method = pass_manager_config.init_method or "default"
     layout_method = pass_manager_config.layout_method or "default"
-    routing_method = pass_manager_config.routing_method or "stochastic"
+    routing_method = pass_manager_config.routing_method or "sabre"
     translation_method = pass_manager_config.translation_method or "translator"
     optimization_method = pass_manager_config.optimization_method or "default"
     scheduling_method = pass_manager_config.scheduling_method or "default"

releasenotes/notes/sabre_level0-1524f01965257f3f.yaml

@@ -0,0 +1,13 @@
+---
+upgrade_transpiler:
+  - |
+    The default routing pass used by optimization level 0 for :func:`.generate_preset_pass_manager`
+    and :func:`.transpile` has been changed from :class:`.StochasticSwap` to :class:`.SabreSwap`.
+    The :class:`.SabreSwap` pass performs exactly the same function but performs better in both
+    runtime and output quality (in number of swap gates and depth) compared to
+    :class:`.StochasticSwap`. For ``optimization_level=0`` this shouldn't matter because it's not
+    expected to run routing for the typical use case of level 0.
+       
+    If you were relying on the previous default routing algorithm for any reason you can use the
+    ``routing_method`` argument for :func:`.transpile` and :func:`.generate_preset_pass_manager`
+    to ``"stochastic"`` to use the :class:`.StochasticSwap` pass.