IBM Promised a Fault-Tolerant Quantum Computer by 2029. Nobody Knows What That Means.
IBM says it will build a fault-tolerant quantum computer by 2029. It has not said what that means.
The company announced a $1 billion commitment to Anderon, a new standalone subsidiary that would become America's first pure-play quantum foundry — a business fabricating quantum hardware for customers outside IBM itself, not just internal R&D. IBM's newsroom says the subsidiary will deliver a fault-tolerant quantum computer by 2029 for commercial clients, contributing to what it estimates could be $850 billion in global quantum economic value by 2040.
IBM has published qubit counts and error rates for its current machines. It has not published the specific technical threshold its 2029 claim depends on, what error rate would constitute passing, or which workloads would count as proof. Classical computers handle errors with checksums and parity bits — redundancy that keeps calculations accurate however long they run. Quantum error correction is the same idea, except that measuring a quantum state disturbs it, making it far harder to catch errors without disrupting the computation itself. That is the core engineering problem. IBM's 2029 promise does not say which problem it has solved.
Morningstar is not buying the timeline. The research firm says IBM will not see material quantum revenue before the 2030s. Different quantum companies define fault-tolerance differently anyway — there is no industry-wide standard that maps "fault-tolerant by year X" to a verifiable milestone. IBM's announcement sits inside that definitional gap.
The Department of Commerce is taking an equity stake in the deal. On May 21, the agency announced $2 billion in CHIPS Act incentives for nine quantum companies — structured as minority, non-controlling equity positions, not grants. IBM received the largest commitment: $1 billion in cash plus intellectual property and workforce. The DoC has used this structure before: it holds a minority stake in Intel worth roughly $40 billion on paper, a position that has gained value. The model is the same here — if the sector matures, the government participates in the upside.
IBM has already deployed more than 90 quantum systems, more than all other industry players combined, and serves 325 Fortune 500 companies. The company argues the scale of its installed base and client network changes the probability of the 2029 timeline landing. The bull case for the equity stake is the same: scale and client relationships are real; the technical milestone is not yet verifiable.
Two of the nine funded companies have principals with Trump administration ties. D-Wave CEO Emil Michael was the Pentagon's top technology official before taking D-Wave public via SPAC in 2022, Reuters reported. PsiQuantum counts Donald Trump Jr.-backed 1789 Capital and Nvidia's venture arm among its investors, per Reuters, having raised $1 billion from them last year.
The DoC says it is building strategic infrastructure. Morningstar says the revenue timeline does not hold. IBM says $850 billion by 2040. These three statements cannot be reconciled until IBM publishes what "fault-tolerant by 2029" actually means in its own technical specifications — and what evidence would confirm it has achieved that milestone.