Supported number of qubits are not increasing with multiple nodes

[intelligi123]

Informations

• 0.14.0:
• Ubuntu 23.10:

What is the current behavior?

I have a qiskit code which qubits I need to increase, for that reason I am using multiple nodes in order to enhance RAM capacity. With one node, I am able to run circuit with 24 qubits having RAM (15G-CPU+5G-GPU). I have increased total resources by adding another node with same specs. I have now total of 40G memory (30G-CPU+10G-GPU) But still my code is not recognizing this increase in memory and generating below Error which is same as I run it with single process.

mpirun -np 2 -machinefile machinefile.txt python3 12.py
Traceback (most recent call last):
  File "/home/dell/12.py", line 44, in <module>
    job_sim = execute(qc,simulator_gpu,blocking_enable=True, blocking_qubits=block_qubits)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/dist-packages/qiskit/execute_function.py", line 286, in execute
    experiments = transpile(
                  ^^^^^^^^^^
  File "/usr/local/lib/python3.11/dist-packages/qiskit/compiler/transpiler.py", line 335, in transpile
    _check_circuits_coupling_map(circuits, coupling_map, backend)
  File "/usr/local/lib/python3.11/dist-packages/qiskit/compiler/transpiler.py", line 458, in _check_circuits_coupling_map
    raise TranspilerError(
qiskit.transpiler.exceptions.TranspilerError: 'Number of qubits (36) in circuit-158 is greater than maximum (29) in the coupling_map'
Traceback (most recent call last):
  File "/home/dell/12.py", line 44, in <module>
    job_sim = execute(qc,simulator_gpu,blocking_enable=True, blocking_qubits=block_qubits)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/dell/.local/lib/python3.11/site-packages/qiskit/execute_function.py", line 286, in execute
    experiments = transpile(
                  ^^^^^^^^^^
  File "/home/dell/.local/lib/python3.11/site-packages/qiskit/compiler/transpiler.py", line 335, in transpile
    _check_circuits_coupling_map(circuits, coupling_map, backend)
  File "/home/dell/.local/lib/python3.11/site-packages/qiskit/compiler/transpiler.py", line 458, in _check_circuits_coupling_map
    raise TranspilerError(
qiskit.transpiler.exceptions.TranspilerError: 'Number of qubits (36) in circuit-158 is greater than maximum (29) in the coupling_map'
--------------------------------------------------------------------------
Primary job  terminated normally, but 1 process returned
a non-zero exit code. Per user-direction, the job has been aborted.
--------------------------------------------------------------------------
--------------------------------------------------------------------------
mpirun detected that one or more processes exited with non-zero status, thus causing
the job to be terminated. The first process to do so was:

  Process name: [[38546,1],0]
  Exit code:    1

Here is my code

from qiskit import QuantumCircuit, QuantumRegister, ClassicalRegister, execute
from qiskit_aer import *

from qiskit_algorithms.utils import algorithm_globals
algorithm_globals.random_seed = 1

def Round_constant_XOR(circuit, source, target, length):
    for i in range(length):
        if (source >> i & 1):
            circuit.x(target[i])
  
def XOR_operator(circ,x,y,result,length):
    for i in range (int(length/2)):
        circ.cx(x[i],result[i])
        circ.cx(y[i],result[i])            
        
def Encryption(qc,a,b,res,length):
    XOR_operator(qc,a,b,res,length)

message = 0x276
key = 0x399
length = 12
ciphertext=""
plaintext = ""

plain_reg=QuantumRegister(length)
key_reg = QuantumRegister(length)
res_enc=QuantumRegister(length)
output=ClassicalRegister(length)

qc = QuantumCircuit(plain_reg , key_reg, res_enc,output)

Round_constant_XOR(qc,message,plain_reg,length)
Round_constant_XOR(qc,key,key_reg,length)

Encryption(qc,plain_reg,key_reg,res_enc,length)

qc.measure(res_enc, output) 

simulator_gpu = AerSimulator(method='statevector', device='GPU')

block_qubits = qc.num_qubits-5
job_sim = execute(qc,simulator_gpu,blocking_enable=True, blocking_qubits=block_qubits)
result_sim = job_sim.result()

print(result_sim.to_dict())
print(result_sim.get_counts())

Steps to reproduce the problem

Running code with mpirun -np 2 -quiet python3 code.py

What is the expected behavior?

It should run faster over multiple processes.

Suggested solutions

It works fine again when I decrease blocking_qubits

---

What is the expected behavior?

Qiskit should recognize that available resources are now increased and it can divide statevector onto two nodes.

---

What is the expected enhancement?

[doichanj]

The statevector simulator requires 16*2^qubits bytes of memory, so you can run up to 30 qubits on 30GB of memory. If you want to add 1 qubit, you have to double the number of MPI nodes