NVIDIA said this on 2025-01-07. The Ledger scores it on 2040-01-06, against criteria fixed on 2026-08-27.
Tracking
We're probably five or six orders of magnitudes away, 15 years for useful quantum computers and that would be on the early side. 30 years is probably on the late side. If you picked 20 years a whole bunch of us would believe it.
Q&A remarks at CES 2025, reported, NVIDIA, 2025-01-07
A genuinely hedged range, not a point claim, so it is tracked at both ends rather than forced into one deadline. EARLY-END check (this claim's primary resolves_by, 15 years out): if a quantum computer has, by 6 Jan 2040, solved a problem of real commercial or scientific value faster or cheaper than the best available classical alternative -- independently verified, not a vendor's own benchmark -- that falsifies even Huang's optimistic bound (NOT_VERIFIED: he was wrong on the early side too). If no such result exists by then, log a check-in and continue tracking toward the LATE-END gate, 6 Jan 2055 (30 years): if by then still no such result exists, that falsifies the claim from the other direction (NOT_VERIFIED: even the generous bound was too generous). VERIFIED only if the first genuine instance lands inside the window itself, closest to the ~20-year midpoint he named as the figure most people in the field would actually believe.
Sources that will settle it: https://thequantuminsider.com, https://www.nature.com, https://www.science.org.
Open this claim in the Ledger, with every other claim and the scorecards. How the Ledger judges a claim.