Advanced Integrated Development Environment for Quantum Computing

AIDE-QC: Software Stack for Quantum-Classical ComputingGetting Started
AIDE-QC is a next-generation software stack enabling heterogeneous quantum-classical programming, compilation, and execution on both near-term and future fault-tolerant quantum computers. Our approach treats quantum computers as co-processors and puts forward single-source C++ and Pythonic programming models for quantum code expression and compilation to native backend gate sets.

AIDE-QC builds upon the service-oriented XACC quantum programming framework and puts forward plugins
for quantum language parsing, intermediate representations, transformations on compiled circuits, error mitigation strategies,
and backend execution and emulation, to name a few. These plugin interfaces enable AIDE-QC to remain flexible as the quantum
computing research landscape grows and advances.

Ultimately, AIDE-QC puts forward a novel C++ language extension for heterogeneous quantum-classical computing called
QCOR . This extension enables programmers to work in C++ and define quantum code as stand-alone
functions or quantum kernels.

Install the AIDE-QC IDE (any OS, requires Docker)

$ /bin/bash -c "$(curl -fsSL"
$ aide-qc --install
$ aide-qc --start my_first_quantum_ide
# IDE will open in browser, ready for work

Install the binaries locally (Mac, Linux)

/bin/bash -c "$(curl -fsSL"

See more details on installation at Getting Started .

Quick Look - Programming Grover’s Algorithm

AIDE-QC promotes a single-source programming model for quantum computing. While most approaches promote circuit construction data structures for remote submission APIs, we enable a true quantum-classical programming language via extensions to existing familiar programming languages, like C++ and Python.

Here we demonstrate how to program a textbook quantum algorithm - the Grover search. We want to implement the general circuit
shown to the right, with initialization, iterative oracle and amplification application, and final qubit measurement. We decompose
the algorithm into 2 library header files and a third implementation file with main() entrypoint.


__qpu__ void amplification(qreg q) {
  compute {
  } action {
    // we have N qubits, get the first N-1 as the 
    // ctrl qubits, and the last one for the 
    // ctrl-ctrl-...-ctrl-z operation qubit
    auto ctrl_qubits = q.head(q.size()-1);
    auto last_qubit = q.tail();
    Z::ctrl(ctrl_qubits, last_qubit);

#include "amplification.hpp"
using GroverPhaseOracle = KernelSignature<qreg>;

__qpu__ void run_grover(qreg q, GroverPhaseOracle oracle,
                        const int iterations) {
  // Put them all in a superposition
  // Iteratively apply the oracle then reflect
  for (int i = 0; i < iterations; i++) {
  // Measure all qubits
#includle "grover.hpp"

__qpu__ void oracle(qreg q) {
    CZ(q[0], q[2]);
    CZ(q[1], q[2]);

int main(int argc, char** argv) {
    auto q = qalloc(3);
    run_grover(q, oracle, 1);
    for (auto [bits, count] : q.counts()) {
      print(bits, ":", count);

To start, we look at the circuit above and notice a sub-circuit called out as `amplification`. It has a very specific structure - a pattern of Hadamard and X gates on all qubits, followed by a multi-qubit control-Z operation, and ended with another Hadamard/X broadcast operation. AIDE-QC and QCOR express this common pattern (so-called compute-action-uncompute) via a special compute {...} action {...} syntax. Note here also that qubit and sub-qreg extraction is possible on a provided qreg, and all single-qubit gates can be controlled on one or many qubits.The AIDE-QC stack allows one to define quantum kernels that can be parameterized with other quantum callables via the KernelSignature type. This kernel is general for any provided oracle, and demonstrates the utility of the classical language extension, whereby all existing classical control flow structures are usable (e.g. for loops).Finally, to use this general grover library code, we just include it as one would for any external library. We define an oracle quantum kernel, and pass it to the general grover call. Our oracle in this example marks states |101> and |011>, so our results should see these states each with 50% probability.
# Compile with qcor, target any quantum coprocessor
$ qcor -qpu ibm:ibmq_vigo -shots 8192 run_grover.cpp
# Execute the binary
$ ./a.out