Occasionally, a single metric quietly changes the direction of an entire field. In December 2025, that metric was 99.99%.
A Sydney-based team at Silicon Quantum Computing (SQC) reported a quantum gate fidelity of 99.99%, a result published in Nature that immediately resonated across the global quantum community. On paper, it appears to be a benchmark. In practice, it marks something far more critical: silicon-based quantum computing crossing from experimental promise into industrial-grade viability.
In the race toward useful, fault-tolerant quantum computers, fidelity isn’t a nice-to-have. It’s the whole game. Quantum systems don’t fail catastrophically. They fail quietly. Tiny errors accumulate, decoherence creeps in, and computations drift off course long before anyone notices. Fidelity is the measure of how faithfully a quantum operation does what it’s supposed to do in an environment that is constantly working against it.
Pushing that accuracy higher has been the field’s defining engineering challenge.
For context, 99% fidelity is often cited as the theoretical threshold for fault tolerance. At that level, error-correcting codes can begin to keep pace with the rate of incoming errors. But theory comes with a cost. At 99%, the overhead is enormous. You can easily need thousands of physical qubits just to produce a single reliable logical qubit.
99.99%—the “four nines” milestone—changes the math.
At this level of precision, the resources required for error correction drop dramatically. Systems become more efficient. Architects become more realistic. And suddenly, the path from prototype to deployable system doesn’t look quite so distant. In classical computing terms, this is the moment when a technology stops being a lab experiment and starts looking like something you can actually build at scale.
For the SQC team, led by Professor Michelle Simmons, this result wasn’t a fluke. It’s the outcome of a fundamentally different approach to building quantum hardware.
While much of the industry relies on quantum dots, superconducting circuits, or trapped ions, SQC takes a bottom-up, atom-by-atom approach. Using scanning tunneling microscopy (STM) lithography, individual phosphorus atoms are placed into a silicon lattice with a precision of 0.13 nanometers. This isn’t probabilistic fabrication. It’s deterministic engineering.
That level of control delivers three meaningful advantages.
First, stability. By embedding qubits directly inside the silicon lattice, they’re naturally shielded from environmental noise. This translates into longer coherence times and fewer spontaneous errors.
Second, predictability. Every qubit is precisely where it’s meant to be. There’s no jitter, no statistical variation, and no need to compensate for manufacturing randomness.
Third—and this is where things get especially interesting—the system improves as it scales. In most quantum architectures, adding qubits introduces more noise, more cross-talk, and lower overall fidelity. The SQC team demonstrated the opposite. Their multi-register processor showed increasing quality as the qubit count grew, a rare inversion of the usual scaling problem.
When you look across quantum modalities in late 2025, the implications are hard to ignore. Superconducting systems from companies such as Google and IBM offer high gate speeds but require complex cryogenic infrastructure and custom fabrication. Trapped-ion platforms deliver high connectivity and strong fidelities, but scaling them introduces their own physical constraints. Silicon-based quantum computing, especially at this fidelity level, brings something neither approach fully offers: a credible path to industrial scale using existing semiconductor infrastructure.
And that matters, because quantum computing doesn’t exist in isolation. It lives inside supply chains, foundries, workforce pipelines, and capital markets. By aligning quantum hardware with the same silicon manufacturing logic that powered classical computing for decades, SQC is effectively tapping into a $500 billion global semiconductor ecosystem.
This is where the conversation shifts from “quantum advantage” to fault tolerance.
Fault-tolerant quantum computing is the point at which systems can run long, complex algorithms without being derailed by cumulative error. With 99.99% fidelity, the Sydney team has effectively cleared the error floor. The most challenging physics problem is no longer the dominant constraint. Attention can move to modularity, interconnects, and scaling qubit counts into the millions.
And this isn’t happening in a vacuum.
Telstra has already reported reductions in model training times using SQC’s quantum machine learning systems. The Australian Defence Force has deployed a rack-mounted version of the technology for specialized applications. These aren’t mass-market deployments, but they’re meaningful signals that silicon-based quantum systems are beginning to transition from experimental platforms into operational contexts.
The next phase for SQC is scaling beyond its current multi-register 11-qubit core into larger, modular arrays. If successful, it positions Sydney—not Silicon Valley or Zurich—as a serious contender in delivering the world’s first commercially viable, fault-tolerant quantum computer.
In a field crowded with bold claims and incremental progress, four nines stands out. Not because it’s flashy, but because it’s foundational. It’s the kind of number engineers trust, investors understand, and ecosystems can be built around.
Quantum computing doesn’t move forward on hype alone. It advances when precision meets scale. And right now, that convergence is happening—quietly, deliberately, atom by atom—in silicon labs in Sydney.
