Nvidia Beats Estimates Again as Forward Revenue Guidance Fails to Impress Wall Street
Nvidia's latest quarterly beat drew a muted response, with energy investors weighing whether hyperscaler power buildout can match the chip demand trajectory.
Nvidia logged revenues of $96.22 billion in results announced during the week of August 24 (2026-08-24), beating expectations in a quarter where analysts had forecast $46 billion in semiconductor sales for the three months to July. Net profit came in at roughly $60 billion, double the year-ago figure. Three years ago, Nvidia earned $6.2 billion. The pace of that profit expansion has few precedents in the semiconductor sector's history.5,2
Wall Street was not impressed. Bloomberg Intelligence titled its earnings response plainly: Nvidia's sales forecast failed to impress. Forward guidance attracted skepticism rather than relief, even as Nvidia's market capitalization sits above $5 trillion. Bloomberg Intelligence drew a key distinction: the demand commitments Nvidia cites in its outlook reflect its own production planning, not purchase orders placed by customers, a clarification that has complicated the revenue-through-2027 narrative.4,5
Nvidia told investors next fiscal year revenues will rise by a further 70%. On the base of $96.22 billion, that puts the company's own projection well above what analysts had been modeling before earnings. Markets chose to discount it anyway.5
Behind the guidance is hyperscaler capital spending. Amazon, Google, Meta, and Microsoft combined are projected to spend $1.5 trillion building data centers this year and next, infrastructure that will run on Nvidia chips. Projected construction budgets and contracted chip orders are different instruments, and traders have started treating them accordingly.5
For electricity markets, the $1.5 trillion buildout implies a power load story that has not fully materialized in long-term supply contracts. The Economist noted that a shortage of electricity could short-circuit Nvidia's growth trajectory. Grid interconnection queues and permitting timelines operate on schedules independent of semiconductor demand cycles, and that gap is growing more visible to data center developers.2
Capital began rotating toward the power side of this equation months before Nvidia's latest results. Fluence Energy shares closed at $24.16 on May 8 (2026-05-08), up 98.2% in a single week after disclosing master supply agreements with two hyperscalers and a record $5.6 billion backlog. The move showed how energy infrastructure can quickly become a leading expression of AI capital spending when chip valuations stall.1
Smaller names pairing secured power with AI compute have attracted similar premiums. Bitzero secured a $2.6 billion long-term tenant for its largest compute block, built around low-cost power deployed alongside Nvidia's Blackwell chips. IREN Limited trades above a $21.75 billion market cap, TeraWulf above $13 billion, and Cipher Mining above $10 billion, all carrying valuations that embed sustained AI electricity demand growth well into this decade.3
Foreign Policy offered a direct historical comparison to Japan's technology expansion in the 1980s. Nvidia's quarterly profit may have doubled in a year; Japan's semiconductor dominance also looked durable before it wasn't. The magazine acknowledged Nvidia's results are formidable. It questioned whether the end of this cycle, when it comes, will look any different from the last one.5
The uranium ETF URA fell 5.89% in data recorded through August 31 (2026-08-31), with the coal ETF off 2.70% in the same reading. Both had been running on expectations of AI-driven incremental electricity load. A muted market reception to Nvidia's 70% revenue guidance, combined with selling in energy-adjacent ETFs, suggests traders are recalibrating the pace of the AI electricity buildout, not abandoning it — but the timeline is becoming a contested number.5,3