Nvidia Extends AI Chip Sales Outlook to $1 Trillion+ Through 2027 at GTC 2026

Silicon roadmaps rarely double guidance in a single year; the closest parallel is Intel’s 1993 Pentium launch, which promised – and briefly delivered – a step-function jump in desktop revenue before competition and physics intervened. Nvidia’s gambit echoes IBM’s 1964 System/360 bet: vertical integration, new instruction sets, and a bet that swelling compute demand will absorb unprecedented capex. The projection rides two longer arcs: (1) AI workloads are devouring ever-larger datacenter budgets, reminiscent of the Moore’s-law-fuelled server boom of 2004-2014, but now with network effects around models, and (2) chipmakers are reclaiming system design as hyperscalers hit energy and bandwidth ceilings. Whether today’s trillion-dollar pledge endures will depend on power-density breakthroughs and geopolitical supply-chain stability—variables that, on a century scale, will shape whether the 2020s AI surge looks like the sustained internet diffusion of the 1990s or the short-lived mainframe glory of the 1950s. Investors should recall that unmet 2011 smartphone forecasts still haunt Intel; history rewards skepticism toward straight-line extrapolations.