Going from small to large-scale quantum computers requires overcoming several challenges. One is how to (optically) address each qubit in a large array, without needing a vast number of input/ output connections to the chip? In fact, quantum chips need to be controlled just like conventional CPUs, which contain billions of transistors, but are interfaced (to a motherboard) using only a few hundred input/ output connections. Achieving this means not only i) manufacturing quantum chips using the same processes used to make conventional CPUs, but also ii) designing the electronic circuits in a way that they can function at the ultralow temperatures (-273 degrees) (required for stable qubit operation). UK-based Quantum Motion (QM) has demonstrated both achievements with its “Bloomsbury” chip (announced at the IEEE International Conference on Electronic Circuits & Systems in 2022), which “lays the foundation for mass producing quantum chips using existing silicon manufacturing processes”. “The big breakthrough is (the proof) that it is possible to create a stable qubit on a standard silicon chip vs. realizing it with (exotic tech) such as superconductors or individually trapped atoms”. Bloomsbury is a 3x3mm2 chip based on QM’s “silicon spin-based qubit architecture”, manufactured by CEA Leti (a large microelectronics research institute in France) using standard 300mm wafer production processes (used for high-yield/ volume chip manufacture). Unlike regular computer chips, Bloomsbury contains thousands of quantum dots into which single electrons can be loaded, one by one, to serve as qubits. “We created bespoke ‘quantum primitives’, our version of the transistor, which we use to trap individual electrons” (i.e. the team was able to isolate & measure the quantum state of a single electron for a period of 9 seconds). “Integrating these on-chip with conventional electronics, which we designed to work at deep cryogenic temperatures, allowed us to read out thousands of quantum devices with only 9 input/ output connections – which has removed a huge bottleneck in scaling”. In a huge leap forward for the mass characterization of such devices, QM demonstrated how 1,024 quantum dots occupying an area less than 0.1mm2 can be measured in 12 minutes (“this is 100x faster than other industry efforts, which can take 24 hours or longer to read the equivalent number of dots”). (Fun fact: the chip’s name references QM’s original HQ in Bloomsbury, working closely with the cryogenic laboratory facilities at UCL, where Virgina Ciriano Tejel – then a PhD at UCL/ today working at QM – was instrumental to the breakthrough). It’s worth noting here, that the quantum industry (in general) has managed to create computers with upwards of 100 qubits, but this orders of magnitude away from the millions of qubits required to drive true (business) value! “It has taken 70 years for transistor development to reach where we are today in computing…we can’t spend another 70 years trying to invent new manufacturing processes to build quantum computers…we need millions of qubits & an ultra-scalable architecture to build them now… our discovery provides us the blueprint to shortcut our way to industrial-scale quantum chip production” (which makes it possible to manufacture a functional, “fault-tolerant” quantum computer, in a reasonable size, at scale). QM was spun out of UCL & Oxford University in 2017 by professors John Morton & Simon Benjamin, and today led by James Palles-Dimmock (CEO). Funds from the new “record-setting round”, will be used to grow the team from 25 to 50 by early 2024 & “work closely with tier-one foundries on further validation processes”. It comes amidst a flurry of deals in the space e.g IQM (FI) (€234m), Pasqal (FR) (€125m), Quantum Machines (IE) (€142m), Alice&Bob (FR) (€33m) & Oxford Quantum Circuits (UK) (€17m), with the global (quantum) market set to grow from $412m in 2020 to $8.6bn by 2027. <Source: techcrunch.com, tech.eu, bosch-presse.de, information-age.com>
