Different qubit platforms are being explored as candidates to become the ‘transistor for the quantum era’. At the end of each year, we evaluate the results produced by researchers and companies with state-of-the-art qubits. We like to plot their performance along two axes, the error rate per qubit (quality) and number of qubits (quantity). The leading platforms in 2024 are superconducting, silicon spin, trapped ions and neutral atoms.
The figure below shows a summary of resultsup until December 2024, as published by researchers and commercial companies. It indicates the area in which results could be produced by classical simulations (blue) and which devices are in the Noisy-Intermediate Scale Quantum class (yellow). The end goal for programmable quantum computers, is to be in the Universal Fault-tolerant regime (green).
Summary of results produced by researchers and companies with state-of-the-art qubits up until December 2024
Companies, research and technology organizations and academic groups are using different methods to push towards fault-tolerant quantum computers. We strongly encourage the quantum industry to publish realistic device development roadmaps, so that the global value chain can align with their requirements.
This year, multiple companies have published development roadmaps for superconducting qubit platforms. Understandably, many keep the ‘secret sauce’ to building a quantum computer at scale to themselves. However, we can carefully draw a few conclusions from looking at the roadmaps of IBM Quantum, Google Quantum AI, Rigetti, IQM and Alice & Bob.
- Devices used for development of utility-scale systems contain more than 100 qubits per quantum chip.
- A modular architecture is used, where chiplets are connected. This way, each chiplet with more than 100 high-quality qubits has acceptable yields in the fabrication process.
- The timeline to achieve useful quantum computing in 2030is set by several companies.
Highlighting two roadmaps as an example:
|
|
||||||||||||||||||||
