Build or Buy your test equipment?

Starting research activities into quantum phenomena is a strenuous task. The upwind for quantum computing in the recent years has nonetheless motivated many newcomers to take on the challenge, be it academic labs, quantum startups or established tech companies. Part of the struggle is that it used to be common practice to design, procure, assemble, and install equipment in-house, as well as program the all software from the drivers up. And for good reasons this is often still the case. At the end of the day it is of most value to the operators themselves to know exactly how their systems work and to be able to fine-tune them to specific requirements while remaining capital-efficient.

However, the field is accelerating and at a certain scale of quantum chip research building a full system based on off-the-shelf components is not trivial anymore. Commercial and academic entities face the same question: how do I most efficiently employ the available resources and what activities are better outsourced in order to progress on my roadmap?

The increased interest in building larger scale quantum computers and chips has sparked demand for integrated solutions. Specifically for cryogenic experiments and solid-state qubit testing, full systems become available on the market. The build-or-buy question is now decisive for the speed in which engineers can carry out the necessary research and development.

Build or Buy in comparison

Build or Buy? Picture courtesy to Softwebsolutions.

To understand why and to whom this build-or-buy question is relevant, we need to understand what has changed in quantum research over the past years. For maturing qubits modalities such as superconducting circuits we see qubit counts to reach the deca- and even hecto-scale. As a result the demands on cryogenic setups, control electronics and classical software are scaling almost superlinearly. Additionally, we observe a shift towards application-oriented research such as running smale-scale circuits, error-correction experiments and fabrication optimization. Both trends render the in-house building of the full experimental setup more difficult and distracting. Outsourcing decisions are generally evaluated based on the distinction between core activities that are critical for the tech roadmap and others which are supportive.

While the build-or-buy decision can be ambiguous for academic researchers, it becomes much clearer for the emerging quantum chip industry. The key challenges of such companies are often seen to be improving the qubit chip and quantum computing stack. Thus, spending time on designing, building, maintaining and operating test setups takes resources away from the the core task. Importantly, a testing system becomes very different from a full-stack quantum computer with converging architectures and larger chips. At what exact point technical and cost demands between a dedicated test setup and a fully controlled computing stack are sufficiently different, needs to be determined from case to case.

At OrangeQS we identified early on that building test equipment for quantum chips is a “heavy-lifting” exercise for chip manufacturers, full-stack integrators and large-scale laboratories. The primary value proposition of our turn-key and fully automated products is therefore to increase the testing throughput, while drastically lowering the cost per tested chip. Additionally, scientists and engineers become available to spent time on interpreting the results rather than troubleshooting the system. Our first generation utility-scale test system will achieve this and enable our launching customer to iterate faster and focus more efficiently on making better qubit chips.