Nvidia Q2 Preview: Blackwell's Peak, Rubin's Debut, China's Return — What Wall Street Really Wants To Hear

As Wall Street bets on a 97% revenue jump to $92.3 billion, all eyes are on Nvidia's blockbuster Blackwell run, the Vera Rubin sequel already in the works, and a possible China comeback for H200 chips.

Advertisement
Read Time: 3 mins
Quick Read
Summary is AI-generated, newsroom-reviewed
  • Nvidia's Q2FY27 revenue is forecasted to nearly double to $92.3 billion year-on-year
  • Blackwell GPU demand remains strong, driving most of Q1 data centre revenue
  • Vera Rubin shipments begin in Q3, targeting agentic AI and advanced workloads
Did our AI summary help?
Let us know.

Nvidia's earnings day has become tech's own version of an Avengers roll call — except instead of superheroes, it's Blackwell, Vera Rubin, and a $92 billion revenue number all assembling for one big showdown. Investors get their answer on August 26, 2026, when the Jensen Huang-led chipmaker reports Q2FY27 results, with experts expecting it to blow past last year's numbers by a wide margin.

Revenue is projected to jump 97% year-on-year to $92.3 billion from $46.7 billion, per Bloomberg's analyst consensus. Adjusted EPS is expected to nearly double to $2.09 from $1.05.

Advertisement

Data centre revenue is pegged at $85.7 billion, more than double last year's $41 billion — split roughly between $43.5 billion from hyperscalers and $41.9 billion from AI Clouds, Industrial, and Enterprise (ACIE). Adjusted net income is estimated at $52.3 billion, up 98% from $26.4 billion, with margins ticking up slightly to 55.35%.

Beyond The Big Billions

Beyond the headline numbers, watch for commentary on Blackwell demand, Vera Rubin's rollout, and progress on the H200 China overhang.

Blackwell remains central to the story: the fastest product ramp in Nvidia's history, with hyperscalers and frontier labs deploying hundreds of thousands of GPUs, per Q1 commentary. It drove the bulk of Q1 data centre revenue, and models like OpenAI's ChatGPT 5.5 were co-designed and trained on it.

Advertisement

ALSO READ: Explained: What Are AI Neoclouds And How Nvidia's Revenue-Sharing Model Has Fueled Its Rise

But last quarter's call pivoted toward what's next: Vera Rubin. Named after the astronomer, it's a next-gen, multi-rack architecture built for agentic AI and reasoning-heavy workloads. CFO Colette Kress said production shipments begin in Q3. CEO Huang went further, predicting Rubin will outdo even Blackwell — "every single frontier model company will jump on Vera Rubin from the get go," he said, calling it off to a "tremendous start."

Finally, watch for updates on H200 access to China. The Financial Times reported last week that Nvidia's H200 chips cleared small-batch entry into the Chinese mainland, with ByteDance and Tencent each receiving roughly 10,000 units.

Advertisement

If Blackwell and Vera Rubin were box office contenders, this wouldn't be a solo release — it'd be a franchise clash, Dune versus Avengers, two juggernauts competing for the same weekend. Except in Nvidia's case, there's no rivalry: Blackwell is still selling out the house while Vera Rubin's trailer is already breaking the internet. For Wall Street, that's the real headline — not one blockbuster, but two, back to back.

ALSO READ: How India Built A $9 Billion Space Economy, And ISRO's Long Flight Plan To 2035

Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.


Loading...