Shixin Zhang

The locality is very important for MPSCircuit. If a non-local two-qubit gate is applied, it will be firstly transformed into a series of local swap gates + local two-qubit gate,...

for unjit version, if AD is not required in your workflow, maybe numpy backend is the fastest

Another possible workaround is when your circuit has some time periodicity, then a scan wrapper can greatly reduce the jitting time, see an example for `Circuit`: https://github.com/tencent-quantum-lab/tensorcircuit/blob/master/examples/hea_scan_jit_acc.py. I believe the...

> Do you think it is doable to speed up the complication for MPSCircuit? Firstly, from physics perspective, a TEBD like algorithm applied on non-periodic and very structured circuit often...

And what is the target circuit metric (qubits number, error, circuit depth, gate number etc.) in your case? Also, have you tried tf backend? the jitting time is much shorter

> The target circuit contains about 80 qubits, and about 60 layers of all-to-all connected non-local 2-qubit gates, making it roughly 3000 non-local 2-qubit gates. To me, the scale for...

can be reproduced, may be due to the same issue for QR and SVD. These operations might not support vmap. **Updated:** nope, jax can vmap qr and svd, the reason...

tf backend vmap is ok but with very low CPU utilization, only around 150% for my test example

For the list of functionalities that a backend framework has to support, try ``[s for s in dir(tc.backend) if not s.startswith("_")]``. The critical operations include the support for complex valued...

Yes, they need to support AD. For numerically stability, we can further customize their AD rules, see https://github.com/tencent-quantum-lab/tensorcircuit/blob/master/tensorcircuit/backends/pytorch_ops.py