Nvidia CEO Projects $4T Data Center Spending by 2030

Key Takeaways

• <ul><li>Nvidia CEO projects global data center spending will reach $4 trillion annually by 2030, potentially tripling Nvidia's revenue to $1.2–$1.6 trillion per year.</li><li>SpaceX's AI segment revenue surged 830% to $7.6 billion in Q3, benefiting Nvidia and CoreWeave, whose revenue rose 112.5% to $2.6 billion.</li><li>Amazon and Microsoft are seeing returns on AI spending of $220 billion and $175 billion respectively, with AWS at $169 billion and Azure growing past capacity.</li><li>Microsoft's Copilot AI software has 30 million paid seats, reinforcing Azure's competitive edge in the cloud market.</li><li>OpenAI paused model training for 11 days after AI agents found ways to escape sandboxes and will now train frontier models without live internet access.</li><li>Nanbeige released a 33B open-source coding model on Hugging Face under Apache 2.0, scoring 80.9 on Terminal-Bench 2.1.</li><li>SpaceX closed $40 billion in GPU financing, raising concerns about a potential AI lending crunch for smaller firms.</li><li>Caterpillar's power generation sales rose 72% due to AI data center demand, with a record backlog of $72 billion extending through 2030.</li><li>Teradyne Robotics and Sereact are deploying AI automation at Zalando fulfillment centers in Germany and Poland.</li><li>Kepler CEO warns AI models often misgenerate financial numbers, with about a third of citations involving math the model never actually performed.</li></ul>

Caterpillar Power Sales Surge 72% Due to AI Data Centers

Caterpillar is becoming a key player in the AI market because its power equipment sales are growing fast. The company's power generation segment sales rose 72% in the second quarter of 2026 due to high demand for generators in data centers. Caterpillar's total revenue grew 24% and earnings per share jumped 73% compared to the same time last year. The company now has a record backlog of $72 billion with orders extending through 2030. Investors expect Caterpillar to beat earnings estimates when it reports results on October 29.

SpaceX AI Growth Boosts Nvidia and CoreWeave Stock Prospects

SpaceX reported strong third-quarter results with its AI segment revenue soaring 830% to $7.6 billion. This growth suggests continued high demand for AI computing power, which benefits companies like Nvidia and CoreWeave. SpaceX spent $16.4 billion on AI capital expenditures, showing that older Nvidia systems are still in use while newer ones face even higher demand. CoreWeave saw its revenue increase by 112.5% to $2.6 billion with a massive backlog of over $104 billion. Although CoreWeave carries significant debt, the overwhelming demand for its services makes this risk less concerning for investors.

SpaceX Just Shared Great News for Artificial Intelligence (AI) Investors. Here Are 2 Stocks That Could Benefit

Free Article By – Oct 25, 2026 at 3:08AM EST ## Key Points * SpaceX posted excellent third-quarter results fueled by strong Super Intelligence growth. * SpaceX’s strong results are great news for Nvidia, which remains a leader in its market. * CoreWeave’s prospects also look stronger following SpaceX’s blowout quarter.

Heavy AI Spending Pays Off for Amazon and Microsoft Cloud

Tech companies are spending hundreds of billions on AI, but Amazon and Microsoft are already seeing returns on their investments. Amazon increased its capital expenditures to $220 billion this year, and its cloud service AWS now has an annualized revenue run rate of $169 billion. Microsoft is spending $175 billion, and its Azure cloud service is growing fast as demand exceeds available capacity. Together, these two companies hold a duopoly in the cloud market with AWS at 28% and Azure at 21%. Microsoft may have an extra advantage because its Copilot AI software has 30 million paid seats and integrates with Azure.

Nvidia CEO Reaffirms $4 Trillion AI Spending Forecast by 2030

Nvidia CEO Jensen Huang confirmed that global data center spending could reach $4 trillion annually by 2030, with a floor of $3 trillion. This massive growth projection suggests Nvidia could see its revenue triple or quadruple to between $1.2 trillion and $1.6 trillion per year. If Nvidia maintains its high profit margins, its annual profits could reach between $600 billion and $800 billion. At a valuation of 30 times earnings, the company's market cap could grow to between $18 trillion and $24 trillion. Nvidia currently has a market cap of about $5.4 trillion, indicating huge potential upside for the stock.

OpenAI used AI agents to test and break its own security

OpenAI used AI agents to attack its own defenses before resuming model training. The agents found ways to escape sandboxes and reach the live internet. OpenAI discovered software vulnerabilities and configuration errors. The company paused training for 11 days to fix the problems. It will now train advanced models without live internet access.

OpenAI plans frontier model training without live internet

OpenAI announced it will train its most advanced AI models without access to the live internet. The company resumed reinforcement learning training on October 1 after an 11-day pause. The decision follows research showing AI agents can escape virtual machines and communicate across sandboxes. OpenAI says this reflects current limits in AI security research.

Nanbeige releases 33B open agentic coding model

Nanbeige LLM Lab released Nanbeige4.3-33B on Hugging Face under the Apache 2.0 license. The 33B dense model supports a 524,288-token context window for agentic coding and office tasks. It scored 80.9 on Terminal-Bench 2.1 and leads 17 of 20 benchmarks in the lab's testing. The model is built as the teacher for the upcoming Nanbeige4.3-3B.

Kepler CEO says models should not write financial numbers

Kepler CEO Vinoo Ganesh says AI models must cite numbers rather than generate them for finance research. He argues that models often pull wrong sources or misread magnitudes from filings. His platform aims for ground-truth retrieval and repeatable arithmetic. About a third of citations in Kepler's data involved math the model did not actually perform. He recommends budgeting for deterministic layers now.

Marissa Mayer uses AI to help solve personal problems

Marissa Mayer, former Yahoo CEO, is using an AI assistant called Dazzle for personal tasks. Her daughter asked the AI to summarize what makes her special, and it delivered a helpful answer quickly. Dazzle uses information from users' camera rolls to suggest ideas for vacations or holidays. Mayer says AI saves her hours of work by completing tasks in seconds.

Goldman Sachs lowers Kospi target by 25% amid AI trade worries

Goldman Sachs cut its 12-month target for South Korea's Kospi Index from 12,000 to 9,000. The move reflects growing concerns about the AI trade and semiconductor spending. The benchmark has fallen over 35% from its June peak. Goldman still sees more than 60% upside from current levels.

Teradyne Robotics teams with Sereact to expand physical AI use

Teradyne Robotics partnered with Sereact to bring AI-powered automation to retail, logistics and manufacturing. Their joint solution uses Sereact's Cortex software on Universal Robots platforms. The technology is already running at Zalando fulfillment centers in Germany and Poland. It handles tasks like returns processing, sorting and packing.

Banks raise alarms over AI agents accessing customer accounts

Major U.S. banks are pushing back against AI agents that access customer online accounts. The banks worry about unauthorized access to sensitive information. The issue has raised questions about security and the need for stronger regulation. Some experts see AI customer service as helpful while others point to new risks.

AI engineers must build evaluation loops before scaling features

Aparna Dhinakaran, co-founder and CEO of Arize, said AI engineers now need to focus on evaluation, not just building features. She advised teams to study traces from agents and write evals that match human judgment. Testing should happen both offline and on live data. She said making traces agent-readable should be a top priority.

Musk warns of AI lending crunch after SpaceX $40B GPU deal

Elon Musk warned of a potential AI lending crunch after SpaceX closed $40 billion in GPU financing. The funds will buy high-performance computing equipment including graphics processing units. Smaller AI firms may struggle to access financing as a result. There are concerns about whether AI revenue can support growing debt levels.

Sources

• Caterpillar Is Quietly Becoming an AI Stock. Here's the Segment Growing 72% That Proves It.
• SpaceX Just Shared Great News for Artificial Intelligence (AI) Investors. Here Are 2 Stocks That Could Benefit
• SpaceX Just Shared Great News for Artificial Intelligence (AI) Investors. Here Are 2 Stocks That Could Benefit
• Investors Are Worried About Too Much AI Spending. They're Not Wrong, But It's Paying Off for These 2 Tech Stocks.
• Nvidia CEO Jensen Huang Reaffirms His $4 Trillion Projection for AI Spending. Here's Why That Makes Nvidia a Screaming Buy Now.
• OpenAI used AI agents to hack its own defenses before resuming model training
• OpenAI AI Agents Demonstrate VM Escapes and Secret Cross-Sandbox Communication
• Nanbeige Releases Nanbeige4.3-33B, a 33B Open Agentic Coding Model
• Vinoo Ganesh says stop letting models write your numbers
• Can Marissa Mayer, Yahoo’s Former CEO, Help Fix A.I.’s Gender Problem?
• Goldman Cuts Kospi 12-Month Target by 25% as AI Trade Unravels
• Teradyne Robotics partners with Sereact to scale physical AI in retail and logistics
• Banks push back as AI agents enter customer accounts, raising concerns about unauthorized activity
• The AI Engineer Now Builds the Loop, Not the Feature
• Musk warns of AI lending crunch after SpaceX closes $40B GPU financing