According to Nature Electronics, the breakthrough was pulled off using a bog-standard 45-nanometre semiconductor manufacturing process. The result is a fully integrated electronic–photonic–quantum system that spits out reliable streams of correlated photon pairs, the fundamental currency for future quantum computing, communications and sensing gear.
This means it can be stamped out in commercial chip foundries and if you stick enough of these together and you’re halfway to large-scale quantum networks.
Boston University associate professor of electrical and computer engineering Miloš Popović said: ““Quantum computing, communication, and sensing are on a decades-long path from concept to reality. This is a small step on that path, but an important one, because it shows we can build repeatable, controllable quantum systems in commercial semiconductor foundries.”
Northwestern University electrical engineering professor and quantum optics veteran Prem Kumar said the work was only possible because of an unusual collaboration between electronics, photonics and quantum measurement.
“The kind of interdisciplinary collaboration this work required is exactly what’s needed to move quantum systems from the lab to scalable platforms,” Kumar said.
Quantum tech will need a constant flow of photon pairs to run properly. To deliver that, the researchers cooked up tiny microring resonators on the silicon chip, each about a millimetre square, to generate correlated photons. These devices have also been flagged by Nvidia boss Jensen Huang as critical for scaling AI compute via optical interconnects.
The problem is microring resonators are fussy. Small shifts in temperature or manufacturing tolerances can knock them out of tune, breaking the delicate quantum light generation. So the team embedded photodiodes inside each resonator to monitor their alignment with the incoming laser light. On-chip heaters and control logic then constantly nudge the resonators back into sync in real time.
Northwestern PhD student Anirudh Ramesh, who led the quantum measurements said: “What excites me most is that we embedded the control directly on-chip, stabilising a quantum process in real time. That’s a critical step toward scalable quantum systems.”
Each chip hosts a dozen of these quantum light factories, all running in parallel without stepping on each other’s toes. The self-correcting feedback loops mean the photon sources remain stable even as temperatures drift or neighbouring devices interfere.
Boston University PhD student Imbert Wang, who led the photonic device design said: “A key challenge relative to our previous work was to push photonics design to meet the demanding requirements of quantum optics while remaining within the strict constraints of a commercial CMOS platform. That enabled co-design of the electronics and quantum optics as a unified system.”
The chip was built using the same 45-nanometre CMOS platform that BU, UC Berkeley and Silicon Valley outfit Ayar Labs originally developed with GlobalFoundries. That process is already used for optical interconnect chiplets in AI and supercomputing. By looping in Northwestern, the same manufacturing flow now churns out scalable quantum photonic systems.
UC Berkeley PhD student Daniel Kramnik, who handled the chip design, packaging and integration said: “Our goal was to show that complex quantum photonic systems can be built and stabilised entirely within a CMOS chip. That required tight coordination across domains that don’t usually talk to each other.”
As quantum systems get more ambitious, these chips could become the Lego bricks for secure communication networks, precision sensing and eventually quantum computing infrastructure.
Some of the graduate authors have already jumped to industry heavyweights. Josep Maria Fargas Cabanillas and Anirudh Ramesh are now at photonic quantum computing startup PsiQuantum. Ðorđe Gluhović and Sidney Buchbinder joined Ayar Labs. Imbert Wang landed at Aurora, while Kramnik went to Google X and is exploring his own silicon photonics startup.
The project was backed by the US National Science Foundation’s Future of Semiconductors programme, the Packard Fellowship for Science and Engineering and the Catalyst Foundation. Ayar Labs and GlobalFoundries chipped in for fabrication support.
