Huang’s Nvidia Forecast in Focus at Q3 Earnings

“`html

Jensen Huang attends a reception for the 2025 Queen Elizabeth Prize for Engineering, at St James’ Palace in London, Brirain, Nov. 5, 2025.
Yui Mok | Via Reuters
Nvidia CEO Jensen Huang recently disclosed a staggering $500 billion order backlog for its AI-powering chips, spanning 2025 and 2026. This revelation followed a period where the company’s quarterly revenue has surged nearly 600% over four years, signaling that Nvidia confidently anticipate continued, albeit moderating, growth in its next chip cycle. The implications are that the AI boom is far from over.

“This is how much business is on the books. Half a trillion dollars worth so far,” Huang stated at the company’s GTC conference.

The figures include revenue recognized so far in 2025, projected income from the current Blackwell GPUs, and anticipated sales of the Rubin GPUs slated for next year, along with associated networking components. Analysts interpreting Huang’s statements believe this suggests considerably higher revenue in 2026 than previously estimated by Wall Street consensus.

Wolfe Research analyst Chris Caso noted that Nvidia’s disclosures indicate a clear upside to current consensus estimates. Caso estimates data center sales could exceed previous 2026 calendar-year projections by $60 billion, maintaining an equivalent to a buy rating on the stock.

Despite this optimistic outlook, Nvidia’s stock price remains about 5% below its value on October 28, when Huang made his announcement. This discrepancy reflects ongoing investor debates surrounding the AI boom and concerns about potential overspending on infrastructure by major cloud companies (hyperscalers) and AI research labs.

Ahead of Nvidia’s third-quarter earnings report, analysts expect earnings per share of $1.25 on sales of $54.9 billion, a 56% year-over-year increase. January-quarter guidance is anticipated at $61.44 billion, suggesting a reacceleration of growth.

Nvidia traditionally offers forward-looking guidance for only one quarter. Therefore, any insights Huang provides regarding the sales backlog and outlook for the 2026 calendar year will be carefully examined, not just for implications for Nvidia but also for the broader tech industry. Currently according to LSEG, analysts anticipate Nvidia sales to reach $286.7 billion in 2026.

‘Insatiable AI appetite’

Huang stated at the Washington conference that the company has “visibility” into these revenues. Given Nvidia’s extensive customer base, which includes nearly every major tech company like Google, Amazon, Microsoft, and Meta, this is unsurprising.

In the previous quarter, these hyperscalers all announced increases in capital expenditure allocated to artificial intelligence infrastructure, primarily NVIDIA chips.

Oppenheimer analyst Rick Schafer described capex increases reflect an “insatiable AI appetite,”maintaining a buy rating on Nvidia stock.

Nvidia has actively pursued deal-making during the quarter, and analysts will be interested in details regarding these partnerships.

The largest deal involves Nvidia potentially investing up to $10 billion in OpenAI equity. This deal is purportedly contingent upon OpenAI purchasing between 4 and 5 million GPUs from Nvidia over several years. Nvidia also invested 5$ billion in Intel. The agreement aims to improve the performance of Intel chips when integrated with Nvidia GPUs.

Subsequently, Nvidia acquired a $1 billion stake in Nokia, with intentions of integrating its GPUs into Nokia’s cellular network hardware. Additionally, it continued investing in a wide of startups.

Citi analyst Atif Malik noted that the OpenAI arrangement would be a focal point for investors.

Malik wrote that while there will be concerns about AI capex, supply fundamentally beneath demand. And maintains a buy rating.

Nvidia holds a significant market share, accounting for more than 90% of all AI GPUs. Yet, several of its clients are creating their semiconductors, application-specific integrated circuits, or ASICs, throughout the previous three months for specific uses. The custom chips being produced include Amazon’s Tranium chips, Google’s TPU chips, and OpenAI’s future Broadcom partnership.

Analysts will scrutinize Huang’s commentary on the rising competition from custom silicon and its potential impact on Nvidia’s market position.

These projections exclude sales to China. Earlier in the year, the H20 Chinese chip was prohibited from export, but Huang arranged an agreement to attain export licenses with the government getting 15% of China sales.

Since then, Nvidia’s representatives have given gloomy remarks about sales to China. Schafer, the Oppenheimer analyst, says he thinks China could become an over $50 billion annual prospect for revenue.

Nvidia’s ongoing performance and strategic moves will be essential to defining the development and sustainability of the AI-driven revolution. Specifically, how AI chips are produced from their company.
“`