GroverOperator segfaults on Lightning-Kokkos with +15 wires

[maliasadi]

Issue description

Simulating qml.templates.GroverOperator with Lightning-Kokkos could not be scaled similar to Lighnting-Qubit.

The following code will be killed due to memory segfaults:

import pennylane as qml

num_wires = 15

@qml.qnode(qml.device("lightning.qubit", wires=num_wires))
def circuit():
    qml.templates.GroverOperator(wires=range(num_wires))
    return qml.probs()

circuit()

• System information:

Name: PennyLane
Version: 0.38.0
Required-by: PennyLane-qiskit, PennyLane_Lightning, PennyLane_Lightning_Kokkos, qml_benchmarks

Platform info:           Linux-6.8.0-45-generic-x86_64-with-glibc2.35
Python version:          3.11.8
Numpy version:           1.26.4
Scipy version:           1.14.1
Installed devices:
- lightning.qubit (PennyLane_Lightning-0.39.0.dev40)
- lightning.kokkos (PennyLane_Lightning_Kokkos-0.39.0.dev40)
- default.mixed (PennyLane-0.38.0)
- default.qubit (PennyLane-0.38.0)