Added GHZ demo

- CLI: compile (`-qpu` switch) 

- Execution: state-vector (`qpp`); noisy (`aer`); IBM. Show QObj (IBMQ portal) to demonstate native gate set mapping (e.g. H -> rz and sx decomposition)

- Execution: state-vector (`qpp`); noisy (`aer`); IBM; IonQ. Show QObj (IBMQ portal) to demonstate native gate set mapping (e.g. H -> rz and sx decomposition)

- Advance: increase number of qubits (~50) => MPS simulation (TNQVM); choose IBM pulse-mode (submitting pulses to IBM)

/// Simulator

/// $qcor -qpu qpp ghz.cpp 

/// IBMQ Backend (normal gate-mode)

/// $qcor -qpu ibm:ibmqx2 ghz.cpp

/// IBMQ Backend (pulse mode)

/// QCOR lowers gates --> pulses

/// $qcor -qpu ibm:ibmq_montreal[mode:pulse] ghz.cpp 

__qpu__ void ghz(qreg q) {

  H(q[0]);

  for (int i = 0; i < q.size() - 1; i++) {

    CX(q[i], q[i + 1]);

  }

  Measure(q);

}

int main() {

  set_shots(100);

  auto q = qalloc(3);

  ghz(q);

  auto counts = q.counts();

  for (auto [bit, count] : counts) {

    print(bit, ": ", count);

  }

}

 No newline at end of file