Robert Wille says quantum software needs design automation, not just algorithms

[Sekiro68]

The Quantum Insider has published an interview with Robert Wille, a professor at the Technical University of Munich and co-founder of the Munich Quantum Software Company. He was interviewed by Yuval Boger, chief commercial officer of QuEra. The transcript reportedly dates from early 2025, so some details may have moved on

Wille's central argument is that quantum software has a lot to learn from classical chip design. Compiling a quantum circuit onto hardware, deciding where qubits go and how operations are routed, looks very similar to placement and routing in electronic design automation. Decades of methods from that field could be reused

He expects future systems to be heterogeneous, mixing CPUs, GPUs, quantum processors and AI chips. He also expects more variety in software stacks built on intermediate representations like MLIR and QIR, and supports QDMI as a way to give software hardware agnostic access to devices

On neutral atoms, Wille says compilation involves trade offs that will persist for decades. Placement, moving atoms around, running operations in parallel and choosing error correction codes all interact. There is no single right answer, which makes good tools essential

He describes his own company as open source and an interesting experiment in how to fund that kind of work. It is a refreshing angle, because a lot of quantum attention goes to hardware and algorithms. Without good design tools, neither will reach their potential

Should quantum software borrow more from classical chip design tools? And can open source quantum companies make money?

Robert Wille (Munich Quantum Software Company): Quantum software needs design automation, not just algorithmsthequantuminsider.com