Add Fault-Tolerance Engine end-to-end tutorial

tests/notebooks/test_fault_tolerance_engine_end_to_end.py

@@ -0,0 +1,17 @@
+from tests.utils_for_testbook import wrap_testbook
+from testbook.client import TestbookNotebookClient
+
+
+# The notebook uses the shared public logical noise by default (INITIALIZE_OWN_NOISE is
+# False), so it skips the ~1h logical-noise initialization. It still runs the full backend
+# pipeline (synthesize -> export -> route -> view -> estimate), hence a generous timeout.
+@wrap_testbook("fault_tolerance_engine_end_to_end", timeout_seconds=600)
+def test_notebook(tb: TestbookNotebookClient) -> None:
+    # The circuit was routed onto a real 2-D surface-code patch footprint.
+    assert tb.ref("num_patches") > 0
+
+    # The total logical error is estimated at each requested code distance and
+    # shrinks as the distance grows.
+    total_errors = {int(k): float(v) for k, v in dict(tb.ref("total_errors")).items()}
+    assert set(total_errors) == {11, 13, 15}
+    assert total_errors[11] > total_errors[13] > total_errors[15]

tutorials/technology_demonstrations/fault_tolerance_engine_end_to_end/fault_tolerance_engine_end_to_end.metadata.json

@@ -0,0 +1,10 @@
+{
+  "title": "Fault Tolerance Engine End To End",
+  "subtitle": "End-to-end walkthrough of the Classiq Fault-Tolerance Engine",
+  "description": "An advanced, end-to-end walkthrough of the Classiq Fault-Tolerance Engine: retrieve a public logical noise model, model a QPE circuit in Qmod, synthesize, export to Clifford+T, route it onto a 3-D surface-code lattice, visualize the routed program, and summarize physical-qubit resources, fidelity, and total logical error across code distances.",
+  "friendly_name": "Fault Tolerance Engine End To End",
+  "vertical_tags": [],
+  "problem_domain_tags": [],
+  "qmod_type": [],
+  "level": []
+}