Superconducting Qubits: Build Quantum Computers | QuantumCasts

Scaling quantum computers requires scaling the control and measurement infrastructure: (1) For a 50-qubit system, scaling to 500 qubits requires 10x more RF components and control lines, (2) The dilution refrigerator's cooling power becomes a limiting factor as more components are added, (3) Signal line density becomes a critical constraint. Solutions include: (1) using ribbon cables with higher wire density, (2) cryogenic CMOS (cryo-CMOS) electronics that operate at 4K temperatures inside the refrigerator, (3) integrating control electronics directly on the chip to reduce external components.

Scaling quantum processors requires overcoming significant engineering challenges, including building a talent pipeline by hiring conventional engineers with experience in integrated optics, lithography, solid-state physics, and RF/microwave control, rather than relying solely on quantum physics experts; companies address this by maintaining close university partnerships to recruit skilled graduates and create pathways for bright PhD students to enter the quantum computing industry.

Quantum processors with multiple qubits (such as 16-qubit chips) are wire-bonded to packaging that connects to microwave signal cables. These cables plug into the sides of the chip and route microwave signals to isolators and amplifiers mounted on the same plate. The microwave signals then travel back up through the dilution refrigerator stages to room temperature. Each stage operates at a different temperature, requiring careful routing of cables through thermal shields.

Scaling quantum processors from laboratory to production requires specialized manufacturing facilities. Current quantum processors are often assembled manually, which introduces contamination and reduces qubit coherence times. Scaling to production requires automated manufacturing similar to semiconductor fabrication, with all processes controlled to prevent errors. This transition from laboratory to production is one of the major engineering challenges in quantum computing.

This section addresses the major challenges in scaling quantum computers. Key challenges include: (1) Cryogenic systems are bulky and expensive; (2) Microwave components (attenuators, circulators, cables) are large relative to qubit size; (3) Control electronics at room temperature add noise and require miniaturization; (4) Frequency crowding limits the number of qubits in a given bandwidth; (5) Wiring for million-qubit systems is impractical with current approaches. The speaker emphasizes that miniaturization of microwave components is essential for scaling, and microwave engineering will play a crucial role in achieving breakthroughs.

Noisy Intermediate-Scale Quantum (NISQ) algorithms are hybrid quantum-classical approaches designed to leverage near-term quantum computers with limited qubits (50-100) and high error rates. These algorithms, such as the Variational Quantum Eigensolver (VQE) and Quantum Approximate Optimization Algorithm (QAOA), combine parameterized quantum circuits with classical optimization to solve problems like quantum chemistry simulations and combinatorial optimization. Despite noise and limited coherence times, NISQ algorithms offer potential advantages in energy efficiency and novel solution approaches, though they face challenges including the barren plateau problem, expressibility limitations, and circuit compilation requirements for specific hardware architectures.

Noisy Intermediate-Scale Quantum (NISQ) computers have four key limitations: limited qubits, limited gate depth before decoherence, restricted qubit connectivity, and noisy gates/measurements without built-in error correction. The Variational Quantum Eigensolver (VQE) addresses these constraints as a hybrid quantum-classical algorithm. VQE finds the ground state of a Hamiltonian by preparing a parameterized quantum circuit, measuring the Hamiltonian expectation, and using classical optimization to update parameters iteratively. This approach offloads most computation to classical computers, making it suitable for near-term quantum hardware.

The Variational Quantum Eigensolver (VQE) is a hybrid quantum-classical algorithm designed for near-term noisy intermediate-scale quantum (NISQ) computers, where quantum computers prepare parameterized quantum states and classical computers optimize the parameters to minimize or maximize an objective function such as molecular energy; this approach leverages quantum parallelism to evaluate different Hamiltonian terms simultaneously while mitigating noise through classical optimization, making it a promising pathway toward practical quantum advantage in chemistry and materials science applications before full error correction becomes feasible.

Variational Quantum Eigensolver (VQE) represents a paradigm shift for near-term quantum devices. Unlike QML 1.0 requiring fault tolerance, VQE works with noisy hardware using compact circuits. It applies the Ritz variational principle: prepare a quantum state, measure its energy, and use classical optimization to minimize toward the ground state. This approach pulls noise into the model as a feature, using parameterized gates with classical optimization loops to find minima in quantum subspaces.

The Variational Quantum Eigensolver (VQE) is a hybrid quantum-classical algorithm designed to solve the electronic structure problem on Noisy Intermediate-Scale Quantum (NISQ) devices. Unlike classical methods like Hartree-Fock, DFT, and Coupled Cluster that suffer from computational scaling limitations, VQE uses a parameterized quantum circuit (ansatz) to prepare trial wave functions, then employs classical optimization to minimize the energy expectation value. The algorithm maps the chemical Hamiltonian to Pauli operators using techniques like Jordan-Wigner mapping, enabling energy estimation through measurements. VQE provides a lower-bounded energy estimate (variational principle) and can be adapted for various quantum chemistry applications, though it faces challenges including barren plateaus, measurement overhead, and circuit depth limitations on current hardware.

There are currently at least five distinct physical approaches being pursued. Superconducting qubits (Google and IBM) are tiny electrical circuits made from materials that become superconducting at near absolute zero, with fabrication techniques borrowing from classical chip manufacturing but requiring extreme cooling. Trapped ion qubits (IonQ and Continuum) use individual atoms suspended in electromagnetic fields, with much longer coherence times (sometimes measured in seconds) but slower operation and harder scaling. Photonic quantum computing (SiQuantum) encodes qubits in individual particles of light, with the advantage of potentially room temperature operation and leveraging existing semiconductor manufacturing infrastructure, though photons are hard to make interact with each other. Topological qubits (Microsoft) encode quantum information in exotic quantum states called Majorana zero modes, which are more protected against decoherence. Neutral atom qubits (QuEra) use atoms trapped by laser beams, with the flexibility to rearrange atoms mid-computation.

Trapped ion quantum processors use charged atoms confined in optical traps, offering superior coherence times (up to seconds) and higher gate fidelity than superconductors. Major developers include IonQ, AQT, and Honeywell. Alternative platforms include quantum dots in silicon (leveraging semiconductor expertise), photonic qubits (limited by weak photon-photon interactions), and topological qubits (Microsoft's approach with potential error rate improvements). Each platform trades-offs between coherence, controllability, scalability, and error rates.

Among leading qubit technologies (superconducting loops, silicon quantum dots, topological qubits), trapped ions are considered most mature and robust. They demonstrate all capabilities needed for fault-tolerant quantum computers—from atomic clock-level qubits to networked modular architectures. The fundamental advantage is full connectivity through long-range Coulomb interactions, enabling direct communication between any ions without physical wiring. This architecture provides better optimization with lower overhead and noise compared to nearest-neighbor connectivity in competing systems.

Five prominent qubit realizations exist, each balancing advantages against specific challenges: Spin qubits offer extremely long coherence times (seconds) and silicon compatibility but face scalability and entanglement demonstration limits. Superconducting qubits achieve high gate fidelities (~99.99%) and leverage CMOS-compatible fabrication but require expensive cryogenic infrastructure and face crosstalk challenges. Trapped ion qubits provide exceptional coherence and simplified control but need ultra-high vacuum. Photonic qubits enable long-distance transmission but struggle with entanglement and decoherence. Topological qubits theoretically offer noise resistance but remain unproven experimentally. Each modality requires careful consideration of technical trade-offs between coherence, scalability, fabrication complexity, and operational costs.

Four main quantum computing modalities exist: (1) Neutral atoms - cold atoms trapped in vacuum chambers using laser fields (optical tweezers), scalable to thousands of qubits; (2) Ions - charged particles trapped using electric fields with many electrodes; (3) Photons - light particles traveling through optical circuits on chips, easy to manipulate but don't naturally interact; (4) Superconducting circuits - superconducting loops at millikelvin temperatures, used by Google and IBM. Each modality has different trade-offs between scalability, control, and coherence times.