By Andre Saraiva, UNSW
Ion traps currently have the leading edge in terms of qubit quality, amassing broken records and “world’s-firsts” for qubit fidelity and multiqubit error correction. Recently it has culminated in the first fault tolerant demonstration of a logical qubit. The natural question that comes to mind is “what are the cons?”. And one of the possible answers is the control, which is based on multiple lasers. Light is used for virtually everything in an ion trap quantum computer – ionizing the atoms, cooling them down, controlling the quantum state and finally reading it out. This implies a complicated set of lenses, mirrors, and other optical devices to direct these beams and focalise them on tiny atoms. A packaging nightmare, which requires a mix of science and art to finely tune.
Scientists from MIT have now created a chip that can integrate all necessary light beams and trap the ions monolithically, removing all free space optical equipment. The device is fabricated following the rules of CMOS foundries, which can be leveraged for massive production and for more complex device fabrication. Indeed, semiconductor giant GlobalFoundries is reportedly deeply involved in silicon photonics and routinely fabricates the waveguides and grating couplers required for this type of device.
As a side benefit, the impact of chip vibrations is significantly reduced by this technology. That is because both the laser and the trapped ion move together, such that fluctuations in intensity due to the relative movement are mitigated. The scientists confirmed this advantage by observing the degradation of the qubit coherence when they increase amplitude of vibrations artificially by clamping the trap mount to the vibrating cryocooler. They observed that the integrated photonic control did not lose quality, while a control scheme based on free space lasers decayed significantly as the ion shaking acceleration increased.
So, what is next? While the on-chip integration of control with CMOS technology may sound like a significant progress in the prospects of scalability, it is unclear if in the immediate future the integrated photonics will translate into larger numbers of qubits. That is because the dimensions of the windows that emit the light beams are large, limited by the wavelength of the light needed for the operation of the qubit. The demonstration shows a single qubit in the trap at the centre of the chip. One could imagine having more ions in the same focus and addressing them by frequency modulation. But it becomes unclear whether the solid-state photonic host will impact the fidelity of these control pulses. Moreover, one can only add so many ions to the same trap, and two neighbouring processing units with a trap in each centre would be too far to interact with each other. We might be many years and many inventions away from seeing this CMOS-compatible integration translating into a massively fabricated large-scale quantum computer.
Dr. Saraiva has worked for over a decade providing theoretical solutions to problems in silicon spin quantum computation, as well as other quantum technologies. He works on MOS Silicon Quantum Dot research and commercially-oriented projects at the University of New South Wales (UNSW).
October 30, 2020