Lightning GPU Installation

I’m trying to build using the PyPI/Conda installed cuQuantum package
In the second step, according to the tutorial, it should be python build_ext
But I don’t know which directory should enter the above command.

Hey @Cc1! You should clone the pennylane-lightning repo and run python build_ext within that repo. You should see a file once you clone it, which is what you need to run!

@Cc1 If you are using Conda, you may now install the pennylane-lightning-gpu package available on conda-forge. You can install it with mamba as follows

mamba create -n pll-gpu -c conda-forge pennylane-lightning-gpu

or with conda as follows

conda create -n pll-gpu -c conda-forge pennylane=0.29.0 pennylane-lightning-gpu

If you are using conda, you will likely to specify the pennylane version as above, because conda currently has a problem resolving the right version of pennylane.

Hi @Cc1
If you want to use the package without compilation, you can simply install it via:

python -m pip install cuquantum pennylane pennylane-lightning-gpu

If you are running on ARM or PowerPC you will need to build from source. Let us know if any of the above helps.

Thanks for your help, my problem has been solved.

1 Like

Thanks for your help.

1 Like