search

QuEraComputing / Bloqade.jl

Package for the quantum computation and quantum simulation based on the neutral-atom architecture.
https://queracomputing.github.io/Bloqade.jl/dev/
Other
181 stars 35 forks source link

[Feature Request] GPU kernel for "measure" and "rydberg_corr" #372

Open hxwang-amazon opened 2 years ago

hxwang-amazon commented 2 years ago

Is your feature request related to a problem? Please describe. It seems measure and rydberg_corr functions (I'm not sure if rydberg_density and apply are also the case) are performed on CPU rather than GPU, when running the Bloqade simulation on GPUs (with CUDA and Adapt).

Describe the solution you'd like @Roger-luo suggests writing a GPU kernel for these functions

Describe alternatives you've considered Not using measure in the evolution simulation process.

Additional context

RamAIbot commented 1 year ago

Hi @weinbe58 I would like to work on this issue for unitary hack. Can you provide more details on which function should I parallelize using CUDA ?

weinbe58 commented 1 year ago

@RamAIbot A good place to start would be to look at https://github.com/QuEraComputing/Bloqade.jl/pull/449 for some initial work I did on making a CUDA kernel for measure