[Prototype] Caching in QiskitDevice

pennylane_qiskit/qiskit_device.py

@@ -114,8 +114,8 @@ class QiskitDevice(QubitDevice, abc.ABC):
 
     _eigs = {}
 
-    def __init__(self, wires, provider, backend, shots=1024, **kwargs):
-        super().__init__(wires=wires, shots=shots)
+    def __init__(self, wires, provider, backend, shots=1024, cache=0, **kwargs):
+        super().__init__(wires=wires, shots=shots, cache=cache)
 
         # Keep track if the user specified analytic to be True
         if shots is None and backend not in self._state_backends:
@@ -379,6 +379,20 @@ def analytic_probability(self, wires=None):
 
     def batch_execute(self, circuits):
 
+        # Check if any of the circuit results was cached
+        if self._cache:
+            non_cached_circuits = []
+            cached_results = {}
+            for idx, circuit in enumerate(circuits):
+                circuit_hash = circuit.graph.hash
+                if circuit_hash in self._cache_execute:
+                    cached_results[idx] = self._cache_execute[circuit_hash]
+                else:
+                    non_cached_circuits.append(circuit)
+
+            # Only keep the non-cached circuits
+            circuits = non_cached_circuits
+
         compiled_circuits = []
 
         # Compile each circuit object
@@ -393,9 +407,11 @@ def batch_execute(self, circuits):
             compiled_circ.name = f"circ{len(compiled_circuits)}"
             compiled_circuits.append(compiled_circ)
 
-        # Send the batch of circuit objects using backend.run
-        self._current_job = self.backend.run(compiled_circuits, shots=self.shots, **self.run_args)
-        result = self._current_job.result()
+        if compiled_circuits:
+
+            # Send the batch of circuit objects using backend.run
+            self._current_job = self.backend.run(compiled_circuits, shots=self.shots, **self.run_args)
+            result = self._current_job.result()
 
         # Compute statistics using the state and/or samples
         results = []
@@ -412,6 +428,21 @@ def batch_execute(self, circuits):
             res = np.asarray(res)
             results.append(res)
 
+        if self._cache:
+
+            # Store the computed results if applicable
+            for res, circ in zip(results, circuits):
+                circuit_hash = circ.graph.hash
+                if circuit_hash not in self._cache_execute:
+                    self._cache_execute[circuit_hash] = res
+
+                    if len(self._cache_execute) > self._cache:
+                        self._cache_execute.popitem(last=False)
+
+            # Insert the cached results obtained at the start
+            for k, v in cached_results.items():
+                results.insert(k, v)
+
         if self.tracker.active:
             self.tracker.update(batches=1, batch_len=len(circuits))
             self.tracker.record()