Photo Credit: Masaru Kamikura/Flickr.com

Last week NVIDIA (Nasdaq: NVDA) reported its third quarter results that continued to beat all estimates. NVIDIA continues to doubt critics and analysts who are worried about the bursting of an AI-bubble. According to its predictions, they are seeing “something very different.”

NVIDIA’s Financials

NVIDIA’s revenues grew 62% to $57 billion, ahead of market estimates of $54.9 billion. Non GAAP EPS grew 65% to $1.30, also ahead of market estimates of $1.25.

By segment, data center division, which includes AI chips and related parts, grew 66% to $51.2 billion versus market estimates of $49.1 billion. Its gaming division grew 30% to $4.3 billion, professional visualization segment grew 56% to $760 million, and the automotive and robotics division grew 32% to $592 million.

Due to the geo-political conditions, NVIDIA cannot ship its current-generation Blackwell chips to China. But it got export licenses for the H20 chip. For the current quarter, the company forecast revenues of $65 billion, ahead of analyst estimates of $61.7 billion.

NVIDIA’s Growth Focus

Analysts may be worried about the AI bubble. But NVIDIA is not very concerned. Recent quarterly reports have revealed how companies like Meta are seeing improvement in user metrics due to the deployment of new generative AI capabilities. Meta is just one such example. As more companies migrate their non-AI solutions to AI offerings, they will see the benefits of AI through higher revenues, user engagement, and other metrics.

This trend is not restricted to gen-AI solutions. Organizations are already investing in agentic and robotic AI and are seeing increased efficiencies from these investments. Higher demand for generative, agentic, and robotic AI will translate to higher AI compute requirements, and NVIDIA is well-positioned to address these.

NVIDIA anticipates $3-$4 trillion in annual AI infrastructure spending by the end of the decade. In recent quarters, demand for AI tools has always exceeded their expectations, and NVIDIA is convinced that the trend will continue.

NVIDIA’s New Partnerships

To continue to meet this growing demand, NVIDIA has recently entered into a series of partnerships. Earlier this month, Microsoft, NVIDIA, and Anthropic announced strategic partnerships within AI. The first ever NVIDIA and Anthropic partnership will mean that the two players collaborate on design and engineering with the goal of optimizing Anthropic models. Anthropic’s compute commitment will initially be up to 1 gigawatt of compute capacity with NVIDIA Grace Blackwell and Vera Rubin systems.

Microsoft and Anthropic are also expanding their partnership to provide broader access to Claude for businesses. The partnership will make Claude the only frontier LLM model available on all three of prominent cloud services. Microsoft has also committed to continuing access for Claude across Microsoft’s Copilot family, including GitHub Copilot and Copilot Studio. As part of the partnership, NVIDIA and Microsoft are committing to invest up to $10 billion and up to $5 billion respectively in Anthropic.

Last quarter, OpenAI and NVIDIA also announced a partnership to deploy at least 10 gigawatts of NVIDIA systems for OpenAI’s next-generation AI infrastructure. To support this deployment, NVIDIA will invest up to $100 billion in OpenAI as the new NVIDIA systems are deployed. The first phase is targeted to come online in the second half of 2026 using the NVIDIA Vera Rubin platform.

This quarter, NVIDIA also announced that it was partnering with other industry leaders, including Google Cloud, Microsoft, Oracle and xAI, to build AI infrastructure for the US. These companies are working with the U.S. Department of Energy’s national labs and to build an AI infrastructure that will support discovery and economic growth.

NVIDIA may think that it is at the onset of the AI super cycle, but analysts disagree. While they agree with the benefits that AI provides, they disagree with the notion that the demand will continue at the same pace. A recent McKinsey report points out that the return on investment on the technology is still absent. Also, just 3% of people pay for AI. Analysts expect big tech to pour in $3 trillion on AI infrastructure by 2028. Irrespective of who funds it, there simply won’t be enough demand for the capacity that is being built out. Companies like OpenAI looking to grow their businesses through increased use of AI are also in for a surprise.

Additionally, the market is characterized by heavy loans and odd deals. Companies are investing heavily into AI infrastructure development, but as per Goldman Sachs analysts, these hyperscalers have taken on $121 billion in debt over the past year, a more than 300% uptick from the industry’s typical debt load. If one looks at the recent Meta and Blue Owl deal, it is not very different. The way the deal is structured, while the $27 billion loan doesn’t show up on Meta’s balance sheet, if the data center fails, Meta is still responsible to pay Blue Owl for the value of the data center.

Even the NVIDIA OpenAI deal mentioned above suggests that NVIDIA will bankroll OpenAI’s data centers. OpenAI will use Nvidia’s chips in those data centers, thus implying that Nvidia is subsidizing OpenAI and artificially inflating actual demand for AI.

Even if one assumes that AI demand would far exceed capacity, there is another limiting factor to building capacity – energy. Data centers that fuel AI models account for about 6% of total US electricity demand, and that share is expected to grow to 11% by 2030.

Energy production in the country is not keeping pace with this growing demand. US peak summer spare power generation capacity decreased from 26% five years ago to 19% this year. Setting up renewable power plants takes longer, and with recent political conditions in the country, those plants are not coming up fast enough. It could well be the case that data centers are established, but they have little or no power flowing through them.

The recent sale of stake by Peter Thiel and SoftBank also makes one think that NVIDIA may be over selling the AI hype.

NVIDIA is currently trading at $182.55 with a market capitalization of $4.4 trillion. It hit a 52-week high of $212.19 earlier last month and has soared from the 52-week low of $86.62 in April.

More By This Author:

Cloud Stocks: Analysis Of Palo Alto Networks’ Chronosphere Acquisition
Meta Pours Funds Into AI Capacity Build-Out
Cloud Stocks: Alphabet Proves AI Is Strengthening, Not Replacing Its Core Search Business