Nvidia GTC 2026: Jensen Huang's Bullish Keynote Fails to Move Wall Street
Summary
- • Nvidia stock dropped during Jensen Huang's 2.5-hour GTC keynote on Monday
- • Huang projected $1 trillion in Blackwell and Vera Rubin chip orders by end of 2027
- • Wall Street unmoved despite 73% year-over-year revenue growth last quarter
- • Analysts cite AI uncertainty and ROI ambiguity as reasons for investor caution
Details
Nvidia stock fell during the GTC keynote despite bullish announcements
Investors were unimpressed with Huang's 2.5-hour presentation, reflecting broader Wall Street anxiety about AI's uncertain future and bubble fears rather than any weakness in Nvidia's fundamentals.
Nvidia revenue grew 73% year-over-year last quarter
The company has consistently beaten its quarterly estimates by wide margins, indicating sustained and accelerating demand for its AI chips despite investor nervousness.
Huang projects $1 trillion in Blackwell and Vera Rubin chip orders by end of 2027
This figure covers just two of Nvidia's many product lines, suggesting the company's total addressable market is substantially larger. The projection was one of several major data points from the keynote.
Huang framed the AI agent ecosystem as a $35T market and physical AI/robotics as a $50T market
These forward-looking projections frame Nvidia's long-term opportunity. Wall Street's muted reaction suggests investors are discounting speculative TAM figures in favor of near-term certainty.
NemoClaw open-source project launched in collaboration with OpenClaw's creator
Nvidia launched NemoClaw at GTC, built jointly with the original OpenClaw creator. Huang declared every enterprise needs an OpenClaw strategy, framing it as a key pillar of Nvidia's robotics/agentic push.
Olaf robot demo at GTC highlighted Nvidia's physical AI ambitions — and its limits
Nvidia demonstrated a robotic version of Disney's Olaf character, which had its microphone cut after it began rambling. Analysts noted the demo exposed unresolved 'messy gray areas' on the social and safety side of humanoid robots in public spaces.
Amazon confirmed plans to purchase 1 million Nvidia GPUs for AWS by end of 2027
Reported by Reuters and confirmed at GTC, the Amazon deal is one of the largest single GPU procurement commitments publicly disclosed and underscores hyperscaler reliance on Nvidia infrastructure.
Enterprise AI adoption ROI data lags actual deployment by six or more months
Futurum CEO Daniel Neuman noted that surveys showing low enterprise AI adoption cite data that is months old by the time it is aggregated and published. Analysts speaking directly with enterprises report a more advanced picture of actual deployment.
AI's transformational speed has created a new category of uncertainty markets struggle to price
Neuman argued that AI's rapid pace makes it harder, not easier, for markets to assign value — a paradox where rapid progress breeds investor anxiety rather than confidence.
Market Impact = stock/investor reaction, Stat = quantitative data point, Insight = analyst interpretation, New Tech = new hardware/software announcement, Product Launch = demo or product debut, Partnership = major customer deal, Context = background framing
What This Means
Nvidia remains the dominant infrastructure provider for AI, posting record revenue growth and securing massive customer commitments — yet Wall Street is pricing in uncertainty about whether AI's economic returns will materialize at the scale Jensen Huang describes. The gap between Silicon Valley confidence and investor skepticism is not about Nvidia's current business performance, which is exceptional, but about whether the $35–50 trillion market projections are credible or speculative. For enterprises and hyperscalers, the answer appears to be procurement first, ROI articulation later — a dynamic that may eventually close the credibility gap with investors as adoption data catches up to reality.
Sources
Updates
Added two new details from TechCrunch GTC podcast recap: (1) Nvidia launched NemoClaw, an open-source robotics project co-developed with OpenClaw's original creator, with Huang declaring every enterprise needs an OpenClaw strategy; (2) The Olaf robot demo had its microphone cut after it began rambling, with analysts noting unresolved social/safety challenges for humanoid robots in public spaces. Source count bumped to 2.