Fault-tolerant quantum computing (FTQC) is the long-term goal of the quantum computing industry — machines that can run complex quantum algorithms reliably despite the inherent noisiness of physical qubits. In a fault-tolerant system, quantum error correction codes continuously detect and fix errors faster than they accumulate, allowing computations to proceed for as long as needed. This is the regime required for the most impactful quantum applications, including breaking RSA encryption, simulating complex molecules for drug discovery, and solving large optimization problems.
Reaching fault tolerance requires clearing a demanding threshold: the physical error rate of each qubit must be low enough that error correction codes can successfully suppress errors when scaled up. Google's 2024 demonstration that its surface code performed better with more qubits was the first experimental proof that this threshold can be crossed with real hardware. IBM, Quantinuum, and Microsoft have all laid out roadmaps targeting fault-tolerant operation within the next several years, though timelines remain uncertain.
The resource requirements are substantial. Current estimates suggest that running Shor's algorithm to break a 2048-bit RSA key would require roughly 20 million physical qubits — far beyond today's hundreds. More tractable near-term targets, like simulating small molecules relevant to pharmaceutical research, might require tens of thousands to hundreds of thousands of physical qubits. The race to fault-tolerant quantum computing is driving massive investments in qubit quality, error correction efficiency, and system engineering across the industry. For deeper coverage, see DeepTechIntel's quantum computing section.