Quantum error correction is emerging as the central engineering challenge shaping how the quantum computing industry plans, invests, and measures progress as the race toward utility-scale machines intensifies. The push spans multiple fronts—from open-source efforts to share full-stack designs, to corporate roadmaps for larger superconducting systems, and new reports showing quantum approaches can speed up early-stage drug design workflows.
A 2025 technical study led by Riverlane describes a major shift: error correction is no longer treated mainly as theory, but as the practical “center of gravity” for building reliable quantum computers that can do useful work. At the same time, new models for collaboration are taking shape, including a non-profit that says it has built the world’s first open-source, full-stack quantum computer.
Error correction becomes the bottleneck
The Riverlane-led Quantum Error Correction Report 2025 says real-time quantum error correction is now the industry’s defining engineering hurdle, influencing national strategies, private investment, and company roadmaps. The report says the bottleneck is not only qubits, but also the classical electronics required to process error signals quickly and apply corrections within about one microsecond.
According to the report, the industry is moving away from an era focused on early noisy machines and into a phase driven by full-stack engineering integration, including decoding hardware and control systems. It also highlights the scale of the data challenge, describing potential data rates reaching hundreds of terabytes per second for error-correction workflows.
The report says multiple hardware platforms—including trapped-ion, neutral-atom, and superconducting technologies—have crossed error-correction thresholds, which shifts attention toward system integration and decoding. It also describes growing interest in different error-correction codes, noting surface codes remain mature while attention is rising around quantum LDPC, bosonic, and hybrid approaches.
Open-source full-stack quantum efforts
Researchers at the University of Waterloo’s Institute for Quantum Computing and Faculty of Science are working through Open Quantum Design (OQD), which the group describes as a non-profit behind the world’s first open-source, full-stack quantum computer. The Phys.org report says OQD’s stack spans hardware, the electronic and computing layers that run it, and open software.
The same report says OQD’s machine uses trapped-ion quantum computing, which involves isolating ions in a vacuum and manipulating them with lasers and electromagnetic fields so they can function as qubits. OQD was co-founded in 2024 by Crystal Senko, Rajibul Islam, and Roger Melko, alongside CEO Greg Dick, according to the report.
Phys.org also reports that OQD recruits organizational partners rather than onboarding individual users, and says it has agreements with the University of Waterloo, Haiqu, the Unitary Foundation, and Xanadu. The report adds that OQD counts more than 30 software contributors and dozens of lab collaborators, and that the initiative builds on eight years of computing hardware development at IQC.
Fujitsu outlines scaling plans
Fujitsu says quantum computing is positioned to drive major advances across society and business, and it highlights use cases such as materials science, drug discovery, and financial forecasting. Fujitsu also states that it promotes R&D across quantum hardware, foundational software, and practical applications in collaboration with research institutes worldwide.
On hardware milestones, Fujitsu says it jointly developed a “world-leading 256-qubit superconducting quantum computer” with RIKEN in April 2025, and that it has been available to companies and research institutions since the first quarter of fiscal 2025. Fujitsu also says it is developing a 1,000-qubit superconducting quantum computer scheduled for installation in a new building at Fujitsu Technology Park and launch in 2026.
Fujitsu states a medium-term goal to build a superconducting quantum computer exceeding 10,000 qubits by fiscal 2030, and says it is participating in a national project promoted by NEDO that runs until fiscal year 2027. It also says it is exploring diamond spin quantum computing as an alternative approach, and reports achieving what it calls the world’s first high-precision quantum gate operation with an error probability below 0.1% in March 2025.
Quantum annealing shows drug design potential
A Quantum Insider report describes research and product claims around PolarisQB’s Quantum-Aided Drug Design (QuADD) platform, which it says runs on a D-Wave Advantage system with over 5,000 qubits. The report says QuADD uses quantum annealing to generate and optimize drug-like molecular candidates in minutes to hours rather than weeks or months.
In a head-to-head study described in the report using Thrombin as a test case, the article says QuADD generated 3,000 molecules in roughly 30 minutes, while a representative generative AI diffusion model took about 40 hours to generate the same number on a node with a single NVIDIA GPU. The report also says QuADD produced leads with stronger predicted binding affinities and better drug-like properties than the representative AI diffusion model, while emphasizing synthesizability.
The article says QuADD frames molecular discovery as a constrained combinatorial optimization problem, using an approach adapted from the Knapsack Problem and leveraging annealing systems that solve Quadratic Unconstrained Binary Optimization (QUBO) or multi-objective optimization problems. It also states that QuADD can explore a chemical space up to about 10^30 theoretical molecules while optimizing for multiple properties in a few hours.
