Hamiltonian time evolution: which simulator is fastest

Trotterised spin chain: ZZ + XX couplings with local fields

8 simulators ran this circuit at 4 to 24 qubits, 24 workloads in all. TensorCircuit was the fastest in 8, Qulacs in 7, Google qsim in 6 and Qiskit Aer (MPS) in 3. The tables below hold every time as it was measured, with the fastest in each row marked. Open this circuit in the interactive tool to see the hardware that could run it and what a run would cost.

Expectation value, local observable, depth 3

Time per run, by qubit count

QubitsPennyLanePennyLane LightningQiskitQiskit AerQiskit Aer (MPS)Google qsimQulacsTensorCircuit
4 qubits
PennyLane
9.46 ms
PennyLane Lightning
4.43 ms
Qiskit
3.80 ms
Qiskit Aer
2.49 ms
Qiskit Aer (MPS)
2.81 ms
Google qsim
4.11 ms
Qulacs
104 µs
TensorCircuit
267 µs
8 qubits
PennyLane
23.2 ms
PennyLane Lightning
9.09 ms
Qiskit
10.1 ms
Qiskit Aer
5.23 ms
Qiskit Aer (MPS)
5.89 ms
Google qsim
8.72 ms
Qulacs
276 µs
TensorCircuit
468 µs
12 qubits
PennyLane
43.8 ms
PennyLane Lightning
14.5 ms
Qiskit
24.5 ms
Qiskit Aer
13.8 ms
Qiskit Aer (MPS)
9.17 ms
Google qsim
13.3 ms
Qulacs
1.85 ms
TensorCircuit
749 µs
16 qubits
PennyLane
208 ms
PennyLane Lightning
32.9 ms
Qiskit
259 ms
Qiskit Aer
43.3 ms
Qiskit Aer (MPS)
12.3 ms
Google qsim
19.2 ms
Qulacs
34.0 ms
TensorCircuit
4.46 ms
20 qubits
PennyLane
4.57 s
PennyLane Lightning
444 ms
Qiskit
6.72 s
Qiskit Aer
666 ms
Qiskit Aer (MPS)
skip
Google qsim
51.8 ms
Qulacs
711 ms
TensorCircuit
164 ms
24 qubits
PennyLane
skip
PennyLane Lightning
16.1 s
Qiskit
skip
Qiskit Aer
12.9 s
Qiskit Aer (MPS)
skip
Google qsim
1.04 s
Qulacs
21.1 s
TensorCircuit
6.20 s

Expectation value, global observable (all_z), depth 3

Time per run, by qubit count

QubitsPennyLanePennyLane LightningQiskitQiskit AerQiskit Aer (MPS)Google qsimQulacsTensorCircuit
4 qubits
PennyLane
9.82 ms
PennyLane Lightning
4.63 ms
Qiskit
3.88 ms
Qiskit Aer
2.62 ms
Qiskit Aer (MPS)
2.95 ms
Google qsim
4.18 ms
Qulacs
113 µs
TensorCircuit
275 µs
8 qubits
PennyLane
24.4 ms
PennyLane Lightning
9.47 ms
Qiskit
10.6 ms
Qiskit Aer
5.52 ms
Qiskit Aer (MPS)
6.44 ms
Google qsim
8.90 ms
Qulacs
321 µs
TensorCircuit
522 µs
12 qubits
PennyLane
46.3 ms
PennyLane Lightning
14.9 ms
Qiskit
25.7 ms
Qiskit Aer
14.4 ms
Qiskit Aer (MPS)
10.0 ms
Google qsim
13.6 ms
Qulacs
2.51 ms
TensorCircuit
1.42 ms
16 qubits
PennyLane
223 ms
PennyLane Lightning
37.6 ms
Qiskit
262 ms
Qiskit Aer
51.4 ms
Qiskit Aer (MPS)
14.3 ms
Google qsim
19.9 ms
Qulacs
48.2 ms
TensorCircuit
45.6 ms
20 qubits
PennyLane
4.86 s
PennyLane Lightning
563 ms
Qiskit
7.22 s
Qiskit Aer
806 ms
Qiskit Aer (MPS)
skip
Google qsim
69.5 ms
Qulacs
1.02 s
TensorCircuit
1.43 s
24 qubits
PennyLane
skip
PennyLane Lightning
21.9 s
Qiskit
skip
Qiskit Aer
15.9 s
Qiskit Aer (MPS)
skip
Google qsim
1.46 s
Qulacs
26.9 s
TensorCircuit
45.3 s

Expectation value, global observable (pauli_mix), depth 3

Time per run, by qubit count

QubitsPennyLanePennyLane LightningQiskitQiskit AerQiskit Aer (MPS)Google qsimQulacsTensorCircuit
8 qubits
PennyLane
25.0 ms
PennyLane Lightning
9.63 ms
Qiskit
10.7 ms
Qiskit Aer
5.39 ms
Qiskit Aer (MPS)
6.73 ms
Google qsim
9.13 ms
Qulacs
349 µs
TensorCircuit
593 µs
12 qubits
PennyLane
46.9 ms
PennyLane Lightning
15.2 ms
Qiskit
26.0 ms
Qiskit Aer
14.7 ms
Qiskit Aer (MPS)
11.1 ms
Google qsim
14.1 ms
Qulacs
3.17 ms
TensorCircuit
2.12 ms
16 qubits
PennyLane
236 ms
PennyLane Lightning
45.6 ms
Qiskit
271 ms
Qiskit Aer
60.2 ms
Qiskit Aer (MPS)
16.1 ms
Google qsim
23.0 ms
Qulacs
61.5 ms
TensorCircuit
78.8 ms
20 qubits
PennyLane
5.11 s
PennyLane Lightning
795 ms
Qiskit
6.93 s
Qiskit Aer
981 ms
Qiskit Aer (MPS)
skip
Google qsim
158 ms
Qulacs
1.28 s
TensorCircuit
2.65 s

Expectation value, global observable (zz_sum), depth 3

Time per run, by qubit count

QubitsPennyLanePennyLane LightningQiskitQiskit AerQiskit Aer (MPS)Google qsimQulacsTensorCircuit
8 qubits
PennyLane
23.7 ms
PennyLane Lightning
9.40 ms
Qiskit
10.4 ms
Qiskit Aer
5.31 ms
Qiskit Aer (MPS)
6.26 ms
Google qsim
9.07 ms
Qulacs
305 µs
TensorCircuit
537 µs
12 qubits
PennyLane
44.4 ms
PennyLane Lightning
14.8 ms
Qiskit
24.9 ms
Qiskit Aer
14.3 ms
Qiskit Aer (MPS)
10.0 ms
Google qsim
14.0 ms
Qulacs
2.50 ms
TensorCircuit
1.74 ms
16 qubits
PennyLane
211 ms
PennyLane Lightning
38.4 ms
Qiskit
261 ms
Qiskit Aer
51.3 ms
Qiskit Aer (MPS)
13.8 ms
Google qsim
20.7 ms
Qulacs
48.1 ms
TensorCircuit
60.3 ms
20 qubits
PennyLane
4.76 s
PennyLane Lightning
588 ms
Qiskit
6.65 s
Qiskit Aer
808 ms
Qiskit Aer (MPS)
skip
Google qsim
95.0 ms
Qulacs
1.01 s
TensorCircuit
1.90 s

Gradient, local observable, depth 3

Time per run, by qubit count

QubitsPennyLanePennyLane LightningQiskitQiskit AerQiskit Aer (MPS)Google qsimQulacsTensorCircuit
4 qubits
PennyLane
29.5 ms
PennyLane Lightning
8.08 ms
Qiskit
unsupported
Qiskit Aer
unsupported
Qiskit Aer (MPS)
unsupported
Google qsim
unsupported
Qulacs
424 µs
TensorCircuit
442 µs
8 qubits
PennyLane
72.7 ms
PennyLane Lightning
15.9 ms
Qiskit
unsupported
Qiskit Aer
unsupported
Qiskit Aer (MPS)
unsupported
Google qsim
unsupported
Qulacs
1.17 ms
TensorCircuit
962 µs
12 qubits
PennyLane
212 ms
PennyLane Lightning
26.4 ms
Qiskit
unsupported
Qiskit Aer
unsupported
Qiskit Aer (MPS)
unsupported
Google qsim
unsupported
Qulacs
9.19 ms
TensorCircuit
1.61 ms
16 qubits
PennyLane
1.64 s
PennyLane Lightning
116 ms
Qiskit
unsupported
Qiskit Aer
unsupported
Qiskit Aer (MPS)
unsupported
Google qsim
unsupported
Qulacs
181 ms
TensorCircuit
10.7 ms

Which quantum computers can run it

At depth 3, 1,000 shots

Quantum computerCan run it up toBorderline up toEstimated cost at that size
Can run it up to
4 qubits
Borderline up to
32 qubits
Estimated cost at that size
$200.30–$600.90
via AWS Braket
Can run it up to
12 qubits
Borderline up to
32 qubits
Estimated cost at that size
$200.30–$600.90
via AWS Braket
Can run it up to
–
Borderline up to
12 qubits
Estimated cost at that size
$23.80–$71.40
via AWS Braket
Can run it up to
64 qubits
Borderline up to
96 qubits
Estimated cost at that size
Not priced
this vendor publishes no rate card
Can run it up to
48 qubits
Borderline up to
–
Estimated cost at that size
Not priced
this vendor publishes no rate card
Can run it up to
48 qubits
Borderline up to
–
Estimated cost at that size
Not priced
this vendor publishes no rate card
Can run it up to
–
Borderline up to
–
Estimated cost at that size
–
Can run it up to
–
Borderline up to
–
Estimated cost at that size
–
Can run it up to
–
Borderline up to
–
Estimated cost at that size
–
Can run it up to
16 qubits
Borderline up to
96 qubits
Estimated cost at that size
from $0.015
via IBM Quantum. 0.01s at maximum throughput, so this is a floor
Can run it up to
24 qubits
Borderline up to
96 qubits
Estimated cost at that size
from $0.421
via IBM Quantum. 0.26s at maximum throughput, so this is a floor
Can run it up to
16 qubits
Borderline up to
96 qubits
Estimated cost at that size
from $0.421
via IBM Quantum. 0.26s at maximum throughput, so this is a floor
Can run it up to
–
Borderline up to
16 qubits
Estimated cost at that size
$0.725–$2.17
via AWS Braket
Can run it up to
8 qubits
Borderline up to
48 qubits
Estimated cost at that size
$1.90–$5.70
via AWS Braket
Can run it up to
4 qubits
Borderline up to
20 qubits
Estimated cost at that size
$1.75–$5.25
via AWS Braket

Sizes are the grid's own: 4, 8, 12, 16, 20, 24, 32, 48, 64, 96, 128, 156 qubits. A device can run a circuit when a clean run is more likely than not, and is borderline when it is better than one in fifty. The cost is for one run of 1,000 shots at the largest size in the row, as a band and not a quote.

The simulators

Each one across every circuit family

How this was measured

One host, one harness

Every simulator ran the same circuits on one host, single-threaded and core-pinned, driven through one common harness, so any overhead the harness adds is charged to the simulator. Each result was checked against a reference implementation: a simulator that ran fast but returned the wrong answer is left out of the ranking. Nothing is interpolated or rescaled. A time was measured, or it is not shown.

Run 2026-08-10, 8 cores, threads=1. numpy 2.4.6, scipy 1.18.0, torch 2.13.0, numba 0.66.0, pennylane 0.45.1, pennylane_lightning 0.45.0, qiskit 2.5.1, qulacs 0.6.14, tensorcircuit 1.8.0, cudaq 0.15.1. One host and one thread make the comparison reproducible. It is not a prediction for your laptop.

The interactive tool shows the same measurements for any workload you pick, with the quantum hardware that could run it and what a run would cost.

Hardware figures each carry the source they came from and the date they were measured. The oldest in this set is 626 days old. Costs are bands, not quotes.

Host: Linux-6.8.0-1051-aws-x86_64-with-glibc2.35, 8 cores, threads=1, core-pinned. numpy 2.4.6, scipy 1.18.0, torch 2.13.0, numba 0.66.0, pennylane 0.45.1, pennylane_lightning 0.45.0. Every simulator was driven through one common harness, so any harness binding overhead is charged to the simulator rather than hidden.

© 2026 EigenOrbit. All rights reserved.