The round was led by Atomico with additional participation from M12 – Microsoft’s Venture Fund, Founders Fund, Blackrock, Redpoint, Baillie Gifford and others. Earlier funding rounds were led by Playground Global. PsiQuantum is working to develop a one million physical qubit quantum computer based upon a photonic technology that incorporates extensive error correction and can be manufactured in a semiconductor wafer fab.
It’s worthwhile to explain what we know about PsiQuantum’s strategy and how it differs from some of the other quantum hardware companies.
- PsiQuantum will be using a photonic technology. Up until now, this technology has been a larger challenge but a number of companies are pursuing it. It has several potential advantages since photons are less affected by the external environment interactions and do not requiring an expensive dilution refrigerator to supercool the device. Also, photonic technology has the ability to network together multiple photonic chips into a mini quantum internet using photons and fiberoptic cable to create a scalable and expandable system. The biggest challenge with photonic technology historically has been implementing two qubit gates since a photon’s low level of interaction with external influences turns into a drawback.
- PsiQuantum has developed a fabrication process that can be implemented in a standard semiconductor foundry. In fact, they have partnered with GlobalFoundries, one of the major Tier 1 semiconductor foundries to manufacture their chips. This is advantageous because it can leverage the semiconductor industry’s high volume manufacturing experience and could potentially provide economies of scale and an ability to ramp up production faster.
- PsiQuantum intends upon leapfrogging over the NISQ era. Their goal is to develop a million physical quantum qubit machine and then use extensive error correction to provide for 100-300 logical qubits. This number of logical qubits should be enough to perform meaningful quantum chemistry computations for commercially relevant problems. PsiQuantum probably won’t have all million qubits on a single chip, but rather implement the system on several chips which are networked together. They say that the total system size would be roughly the size of an office conference room.
Besides concerns about whether PsiQuantum will encounter any unforeseen execution problems in developing their technology, another unknown is what PsiQuantum will use for their software stack and applications support. Providing an effective quantum computing platform requires much more than creating a good chip as we described in one of our previous articles. Since Microsoft is one of PsiQuantum’s investors, perhaps they will partner with Microsoft’s quantum team and use Microsoft’s Q# and Microsoft’s Quantum Development Kit as their primary vehicle for software support.
There is a growing infrastructure of software developers in quantum today and much of their software activities are focused on utilizing NISQ computers using such algorithms as QAOA and VQE. So even though a larger error corrected computer is better, it may not be quite so easy to convert an existing NISQ application that is running on a competitor’s machine.
Nonetheless, PsiQuantum is taking an interesting and potentially breakthrough approach and we are looking forward to hearing more about their developments as they release more information. For more about PsiQuantum and the investment, you can read a blog article on PsiQuantum’s web site here, a Bloomberg Businessweek article here, and articles on Medium from Atomico and M12-Microsoft’s Venture Fund.
April 7, 2020