Multimode entanglement characterization with Presto
Q: Hi Juan Carlos. Can you introduce yourself and explain the specific focus of your research at KTH?
A: I am a Ph.D. student in David Haviland's group at KTH Royal Institute of Technology. My research interests span several areas, including superconducting quantum computing and Continuous Variable (CV) quantum computation. Specifically, I study multipartite entanglement in microwave frequency combs generated by a Josephson parametric amplifier (JPA).
Q: What is the ultimate goal of analyzing these quantum correlations?
A: My main goal is to analyze how the quantum correlations are generated inside the JPA, entangle as many modes as possible, and eventually generate CV cluster states, which are a fundamental resource for measurement-based quantum computation (MBQC).
Q: Before adopting Presto, how did you manage the hardware for upconverting pump tones?
A: I used to use an arbitrary waveform generator (AWG) and a local oscillator (LO). It had significant limitations: the AWG could only generate up to two tones with no phase control, and it had to be locked to the digitizer to ensure coherence. Using multiple instruments was a constant source of incoherence in our system.
Q: How did Presto change that workflow?
A: Presto allows full phase control of the pump tones and easy scalability. We can now both generate and measure as many tones as we need in a completely coherent way, eliminating the struggle of synchronizing multiple instruments.
Q: You work with frequency combs, which involves dealing with multiple frequencies simultaneously. How does Presto handle that?
A: It simplifies things immensely. You don't need to worry about analog mixers, and you only need a few lines of code to generate a frequency comb. You can easily change the measurement bandwidth and comb spacing, and tune all frequencies within Presto's 1 GHz bandwidth.
Q: How was your experience getting up to speed with the Python API?
A: To be fair, everything feels difficult during the first year of a Ph.D., so it was a bit tough at the start! But it became smooth very quickly. The example scripts are quite useful, and the API allows you to do complex things without needing advanced Python skills. For example, it is very intuitive to perform nested sweeps—like a power sweep while doing a frequency sweep and a DC bias sweep simultaneously.
Q: You are in a unique position being located right next to the IMP offices at KTH. How has the support been?
A: I am a bit spoilt! I just need to walk next door and ask. Even though I pass by their office hundreds of times, they always find the time to help me.