JAQSI

During my PHD I had to ran countless of experiments with quantum circuits, some more, some less successfull. Luckily I did not had to submit, all of them, on a real device; instead we can use simulators! They also enable a feature which we naturally don't have on a real device: backpropagation a.k.a. calculating gradients; pretty handy if someone like me wants to do machine learning with these circuits! Anyway, at some point during my PHD I grew quite some frustration with missing features or bugs in a simulator I was using. So I decided to start building a new one! How hard can it be? What first started as little exercise and weekend-project turned out to be actually quite competitive to the existing simulator we were using our experiments. While speed was not our primary goal, the flexibility and speed at which we could now add features and tailor the simulation towards working with quantum Fourier models was definitely precious. Fast-forward, and a few bugs later we're now even on par or ahead compared to most of the other simulators in terms of speed. We recently wrote a paper about our new simulator (which is of course actively maintained). Feel free to checkout our paper on Arxiv.

I'm very happy that we got accepted at WQS26 with an shorter version of this work.