All We Want for Quantum is PICs

Photonic Integrated Circuits (PICs) have already become a mature platform for classical applications such as datacom/telecom, biosensing, and LiDAR. Now, the quantum community is exploring the ways in which PICs can be used. Photonic quantum computer architectures (such as PsiQuantum and Xanadu) have already gone all-in on PICs. Even trapped ion and neutral atom systems are leveraging PICs for improved beam control/routing, and solid-state defect-based qubit systems, such as memQ's approach, are often building technology onto a PIC backbone.

While PICs offer clear advantages to improved cSWAP and manufacturability, there are several unique challenges and opportunities that arise when applying PICs to quantum technologies. The first and foremost challenge is loss. Typical waveguide losses are many orders of magnitude higher than optical fiber losses, and every additional component contributes to the loss of precious photons. This has become a priority for both foundries and design engineers to reduce losses at both the fabrication and design level. It is now increasingly common to see waveguide losses at the dB/m level and edge couplers and modulators with sub-dB loss. These developments are not only critical to quantum PICs but can be fed back to the greater PIC ecosystem and will benefit classical technologies that can always benefit from more optical power.

Next is the need for cryo-compatibility. The environments that quantum systems operate in bring a new challenge to quantum PIC development. Namely, many systems must operate at cryogenic (<20 K) temperatures to reduce thermal noise or to enable the use of key superconductor materials, such as niobium and aluminum, used in transmons or single-photon detectors. Active components that use doped semiconductors suffer from carrier freezeout, and PIC packaging must be able to withstand extreme thermal cycling. To address this, new PIC technologies, such as Pockels effect-based modulators and novel fiber attach technologies, are being developed.

In order to meet these requirements, substantial innovation must be made to existing integrated photonic platforms, such as the heterogeneous integration of new enabling materials onto existing platforms. These changes require significant R&D capital to transition from the laboratory to fabrication, and they also increase manufacturing costs. At this developmental stage, the manufacturing costs for quantum PICs, arguably the most critical metric of classical PICs, are greatly relaxed, as wafer volumes are comparatively low, and a cost-competitive market for quantum products does not yet exist. R&D costs, on the other hand, remain a challenge, especially for startups. However, R&D is being subsidized by government support and could further benefit from shared development between quantum firms with similar needs.

We understand this very well and are in active development of the integrated photonic components necessary for our platform, including cryogenically compatible modulators and switches, quantum frequency converters, and superconducting nanowire single-photon detectors—all on a commercial foundry platform.

The increasing adoption of silicon photonics (see NVIDIA's announcement), driven by advancements in artificial intelligence, is expected to accelerate the development and integration of photonic technologies across various industries. This trend suggests that photonic components and systems will become more prevalent and play a crucial role in shaping the future of technology.