Add performance benchmarks for single qubit randomized benchmarking

benchmarks/randomized_benchmarking.py

@@ -0,0 +1,70 @@
+# Copyright 2022 The Cirq Developers
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     https://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from typing import List
+import functools
+import numpy as np
+import cirq
+from cirq.experiments.qubit_characterizations import _single_qubit_cliffords, _find_inv_matrix
+
+
+def dot(*args: np.ndarray) -> np.ndarray:
+    return functools.reduce(np.dot, args)
+
+
+class SingleQubitRandomizedBenchmarking:
+    """Benchmarks circuit construction time for single qubit randomized benchmarking circuits.
+
+    Given a combination of `depth`, `num_qubits` and `num_circuits`, the benchmark constructs
+    `num_circuits` different circuits, each spanning `num_qubits` and containing `depth` moments.
+    Each moment of the circuit contains a single qubit clifford operation for each qubit.
+
+    Thus, the generated circuits have `depth * num_qubits` single qubit clifford operations.
+    """
+
+    params = [[1, 10, 50, 100, 250, 500, 1000], [100], [20]]
+    param_names = ["depth", "num_qubits", "num_circuits"]
+    timeout = 600  # Change timeout to 10 minutes instead of default 60 seconds.
+
+    def setup(self, *_):
+        self.sq_xz_matrices = np.array(
+            [dot(*[cirq.unitary(c) for c in group]) for group in _single_qubit_cliffords().c1_in_xz]
+        )
+        self.sq_xz_cliffords: List[cirq.Gate] = [
+            cirq.PhasedXZGate.from_matrix(mat) for mat in self.sq_xz_matrices
+        ]
+
+    def _get_op_grid(self, qubits: List[cirq.Qid], depth: int) -> List[List[cirq.Operation]]:
+        op_grid: List[List[cirq.Operation]] = []
+        for q in qubits:
+            gate_ids = list(np.random.choice(len(self.sq_xz_cliffords), depth))
+            idx = _find_inv_matrix(dot(*self.sq_xz_matrices[gate_ids]), self.sq_xz_matrices)
+            op_sequence = [self.sq_xz_cliffords[id].on(q) for id in gate_ids]
+            op_sequence.append(self.sq_xz_cliffords[idx].on(q))
+            op_grid.append(op_sequence)
+        return op_grid
+
+    def time_rb_op_grid_generation(self, depth: int, num_qubits: int, num_circuits: int):
+        qubits = cirq.GridQubit.rect(1, num_qubits)
+        for _ in range(num_circuits):
+            self._get_op_grid(qubits, depth)
+
+    def time_rb_circuit_construction(self, depth: int, num_qubits: int, num_circuits: int):
+        qubits = cirq.GridQubit.rect(1, num_qubits)
+        for _ in range(num_circuits):
+            op_grid = self._get_op_grid(qubits, depth)
+            circuit = cirq.Circuit(
+                [cirq.Moment(ops[d] for ops in op_grid) for d in range(depth + 1)],
+                cirq.Moment(cirq.measure(*qubits)),
+            )