TLDR
- Nvidia CEO Jensen Huang predicts reasoning AI models will consume 1,000 times more computing tokens than current chatbot models
- Company’s new Blackwell Ultra GB300 GPUs launching this year to handle massive inference workloads
- Huang forecasts AI infrastructure spending could hit $1 trillion annually by 2028, with Nvidia capturing most demand
- Nvidia stock trades at P/E ratio of 44.3, representing 26% discount to 10-year average valuation
- Broadcom earnings beat expectations with 60% AI revenue growth forecast, lifting semiconductor sector
Nvidia CEO Jensen Huang just dropped some eye-opening predictions about the future of artificial intelligence computing. Speaking during the company’s fiscal 2026 first quarter earnings call, Huang revealed that next-generation reasoning AI models could require 1,000 times more computing power than current chatbot applications.
$NVDA Q1 2026
"AI inference token generation has surged tenfold in just one year, and as AI agents become mainstream, the demand for AI computing will accelerate." – Jensen Huang
Revenue +69%
*Data Center +73%
*Gaming +42%
*Professional Vis. +19%
*Automotive +72%
EBIT +28%… pic.twitter.com/3beGpMzwGV— Quartr (@Quartr_App) May 28, 2025
The shift represents a fundamental change in how AI systems work. Current AI chatbots like ChatGPT focus on speed, generating quick one-shot responses for users. These systems were revolutionary but often made mistakes or gave incomplete answers.
The new reasoning models take a different approach. They spend time thinking through problems in the background, cleaning up errors before presenting responses to users. This extra processing makes them far more accurate but creates massive computing demands.
Nvidia’s current H100 GPU, built on the Hopper architecture, dominated data center sales in 2023 and 2024. The company designed these chips for both training AI models and running inference workloads.
New Hardware Race Heats Up
Nvidia responded to growing computing needs with its Blackwell architecture. The new design delivers up to 40 times more performance than Hopper chips for inference tasks.
The company’s GB200 GPU NVLink 72 system currently leads the market for reasoning inference workloads. But Nvidia isn’t stopping there. The chip giant plans to ship its even more powerful Blackwell Ultra GB300 GPUs later this year.
Huang emphasized that hardware must keep improving. If reasoning models take too long to generate responses, people simply won’t use them. This creates ongoing pressure for better performance.
The timing couldn’t be better for Nvidia’s business model. The company has built an entire ecosystem around its chips, including networking equipment and CUDA software platform. Once data centers commit to Nvidia’s system, switching becomes expensive and complicated.
Revenue Numbers Tell the Story
Nvidia’s data center business generated $39.1 billion in revenue during the fiscal 2026 first quarter. That represents a 73% jump from the same period last year. Data center sales now make up 89% of Nvidia’s total revenue.
At the company’s GTC conference in March, Huang predicted AI infrastructure spending could surpass $1 trillion annually by 2028. During the recent earnings call, he said Nvidia expects to capture most of that demand.
The scale of this opportunity becomes clearer when looking at token consumption. Huang noted that reasoning models use 1,000 times more tokens than traditional language models. Tokens represent words, punctuation, and symbols that AI systems process.
Stock Valuation Looks Attractive
Despite Nvidia’s $3 trillion market value gain since early 2023, the stock appears reasonably priced by historical standards. Trading at a price-to-earnings ratio of 44.3, shares sit 26% below the company’s 10-year average P/E of 59.8.
Wall Street analysts expect Nvidia to earn $4.28 per share for fiscal 2026. That puts the forward P/E ratio at just 32.1, well below historical norms.
Using current valuation metrics, Nvidia stock would need to rise 38% just to maintain its current P/E ratio. To reach its 10-year average valuation, shares would need to climb 86%.
Broader Chip Sector Shows Strength
Nvidia got additional support Friday from rival Broadcom’s earnings report. The company beat profit and revenue expectations while forecasting 60% AI revenue growth for the current fiscal year.
While Broadcom shares fell 3.8% on what investors saw as modest results, the underlying AI demand trends remain strong. J.P. Morgan analyst Harlan Sur described the earnings as evidence of “sustained strong AI fundamentals.”
Other chip stocks also gained ground Friday. Advanced Micro Devices rose 0.2%, Intel added 0.5%, and Qualcomm climbed 0.6%. Nvidia shares edged up 0.1% to $140.16 in premarket trading.
Broadcom’s AI revenue forecast of 60% growth reinforces Huang’s predictions about exploding demand for computing infrastructure as reasoning models become mainstream.
Stay Ahead of the Market with Benzinga Pro!
Want to trade like a pro? Benzinga Pro gives you the edge you need in today's fast-paced markets. Get real-time news, exclusive insights, and powerful tools trusted by professional traders:
- Breaking market-moving stories before they hit mainstream media
- Live audio squawk for hands-free market updates
- Advanced stock scanner to spot promising trades
- Expert trade ideas and on-demand support
