TensorCircuit benchmark
Timed against 7 other simulators
TensorCircuit 1.8.0 finished 151 of the 175 workloads it was run on and was the fastest of the 8 simulators in 39. At the median it took 2.2× the time of the fastest simulator in the same workload. What it is and how to install it is in its own documentation.
By circuit family
Where it is fast and where it is not
| Circuit family | Workloads finished | Fastest in | Median time against the fastest |
|---|---|---|---|
| Workloads finished 16 of 16 | Fastest in 5 | Median time against the fastest 2.9× | |
| Workloads finished 24 of 24 | Fastest in 8 | Median time against the fastest 1.7× | |
| Workloads finished 24 of 27 | Fastest in 1 | Median time against the fastest 2.6× | |
| Workloads finished 0 of 16 | Fastest in 0 | Median time against the fastest – | |
| Workloads finished 24 of 24 | Fastest in 1 | Median time against the fastest 2.7× | |
| Workloads finished 11 of 12 | Fastest in 4 | Median time against the fastest 1.9× | |
| Workloads finished 16 of 19 | Fastest in 1 | Median time against the fastest 2.8× | |
| Workloads finished 16 of 16 | Fastest in 4 | Median time against the fastest 2.2× | |
| Workloads finished 20 of 21 | Fastest in 15 | Median time against the fastest 1.0× |
TensorCircuit against each other simulator
On the workloads both finished
| Comparison | Workloads both finished | TensorCircuit faster in | The other faster in | Faster at the median |
|---|---|---|---|---|
| Workloads both finished 127 | TensorCircuit faster in 123 | The other faster in 4 | Faster at the median TensorCircuit 12.4× | |
| Workloads both finished 142 | TensorCircuit faster in 91 | The other faster in 51 | Faster at the median TensorCircuit 3.4× | |
| Workloads both finished 104 | TensorCircuit faster in 104 | The other faster in 0 | Faster at the median TensorCircuit 5.6× | |
| Workloads both finished 118 | TensorCircuit faster in 78 | The other faster in 40 | Faster at the median TensorCircuit 2.8× | |
| Workloads both finished 104 | TensorCircuit faster in 78 | The other faster in 26 | Faster at the median TensorCircuit 5.8× | |
| Workloads both finished 118 | TensorCircuit faster in 66 | The other faster in 52 | Faster at the median TensorCircuit 2.9× | |
| Workloads both finished 151 | TensorCircuit faster in 51 | The other faster in 100 | Faster at the median Qulacs 1.3× |
Where it produced no timing
The status the benchmark recorded
A simulator that did not finish a workload keeps the status the benchmark recorded for it, and is left out of that workload's ranking.
| Recorded status | Operation | Workloads |
|---|---|---|
error | Operation Expectation value | Workloads 14 |
unsupported | Operation Statevector | Workloads 6 |
error | Operation Gradient | Workloads 4 |
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.