A (qu)bit off the beaten path
Why we invested in equal1
btov has been active in the quantum technologies (QT) domain for quite a while. For us, QT is a natural extension of our long running activities in electronics and photonics. Our first “proper” QT investment was in quantum sensing player NVision Imaging Technologies in 2017, followed by our engagement with quantum software company HQS Quantum Simulations in 2019.
As trained physicists, we have passionately observed the global quantum computing (QC) community evolve over the last five years — many concepts which were the domain of early theoretical papers and foundational quantum experiments some 25+ years ago (when one of us was doing his PhD) have now been translated into thriving startups. Europe’s extremely strong scientific base in quantum computing did, however, not lead to the first wave of such QC startups: D-Wave, Rigetti, IonQ or PsiQuantum — well known and many of them extremely well-funded — were all founded in the US. But Europe, over the last three years, has very clearly woken up and by now we have an interesting ecosystem of startups covering all quantum computing concepts: the likes of IQM, AQT, Pasqal and Orca have managed to raise respectable amounts of capital and further QC hardware startups with even more advanced ideas (and very much on our radar screen!) will follow suite.
So why would it need yet another QC company?
Firstly, many of the existing concepts (for example from the ion trap or the superconducting domain) have still not given a clear answer as to how to scale to very large (qubit count) sizes — whereas those which have an answer to that (mostly in the photonic domain) all require new manufacturing processes (and yield breakthroughs) which are extremely costly to develop. equal1 is able to use an existing, established CMOS process technology (FD-SOI), which allows them to capitalize on the many €100bns of R&D and Capex that have gone into well-established technology nodes. With equal1 technology, it should be feasible to scale to many hundreds of thousands and eventually millions of qubits — without having to employ new process innovations — and therefore, we believe, at a fraction of the investment needs of competing platforms.
Secondly, equal1 is going to pursue a specialized application field: quantum machine learning and quantum optimization. In that sense, equal1 does not have to pursue the goal of a fully error corrected gate based universal quantum computer. Their technology platform already starts to become useful at much less stringent requirements for qubit lifetimes and gate fidelities. equal1’s computing devices will be used as analog co-processors for extremely compute intensive applications. This should shorten equal1’s time to market from a decade to probably two years. In the ML acceleration application domain they will compete with neuromorphic chip innovators and other deep physics based technologies like photonics based analog computers with an interesting startup ecosystem in itself, which, however, also have to rely on tweaks in existing manufacturing technologies.
And thirdly, we were drawn to this opportunity as it offers substantial energy savings potential for one of the fastest growing power consumers in the global internet infrastructure — deep learning focused server farms in data centers. We believe equal1 has a technology platform which will allow to combine lowest € per qubit and lowest kWh per compute performance metrics — at very compact form factors, fitting under an office desk, and not occupying a whole room of cryogenics. This combination fits perfectly with our combined resource efficiency & pervasive industrial digitalization Fund focus.
equal1 has a unique personnel setup, with highly industrially experienced semiconductor physicists and engineers based in Dublin and in Silicon Valley — using manufacturing facilities here in Germany. Their “let’s get to an MVP — fast” spirit has inspired us; equal1 will demonstrate the next iteration of their technology platform during 2022. But at the same time we are aware that we have quite a journey in front of us. If you are a corporate customer interested in exploring practical large scale QC enabled machine learning or optimization; an engineer or physicist not wedded to the current “high fidelity qubit” orthodoxy; or a fellow investor interested in a cost efficient pathway to practical quantum computing — you know where to find us.
