InvestAnswers|28. Mai
AI Gold Rush Ignites! 🚀 CapEx Boom, IPO Hyper-Acceleration + Chip Rebellion
🚀 AI Gold Rush Ignites: A New Economic Era Begins
This analysis delves into the current, massive shifts in the AI sector, reminiscent of a gold rush. Experts describe this change as unprecedented.
Key Trends and Events:
- IPO Super Cycle: Cerebras' IPO (doubling its price) was the starting signal for an expected $4 trillion IPO cycle. Major upcoming IPOs include SpaceX (in 14 days) and OpenAI (planned for September).
- Compute Arms Race: The hunger for computing power is immense. New model versions (GPT-5.5, Gemini IO) and funding rounds (Anthropic: $30 billion) are driving development.
- Age of AI Agents: The transition to agentic AI and Artificial General Intelligence (AGI) is underway. AI models are already solving complex, previously unsolvable math problems. 2026 is seen as the year of autonomy and the singularity.
- New Threat from XAI: The combination of SpaceX's massive computing power and XAI's rapid iteration poses a serious threat to established players like OpenAI and Anthropic. A proprietary AI training stack built in raw C is up to 10 times faster than Google's framework.
The analysis follows the money to identify the true winners.
AI Revenue:
- Nvidia dominates with $253 billion in AI-related revenue.
- Microsoft, Alphabet, and Amazon follow.
- XAI: Current revenue is small, but a deal with Anthropic ($15 billion/year) could make XAI the #3 player on this list next year.
AI Spending (CapEx):
- Spending on AI infrastructure is historically unprecedented, surpassing past investments in roads, railways, or factories. Goldman Sachs forecasts cumulative investments of $7.6 trillion by 2031.
- The main spenders are Amazon, Google, Microsoft, and Meta. This money ultimately flows to chip manufacturers (Nvidia, Broadcom, AMD, Marvell).
- Strategy: One should invest in assets that are on the receiving end of this massive CapEx boom.
The enormous demand is causing shifts in the chip industry.
- ASIC Rebellion: Major hyperscalers are developing their own custom chips (ASICs) because Nvidia's offerings are insufficient or too expensive.
- Focus on Edge Computing: The future lies in computing power at the network edge (e.g., in cars, robots, devices). Nvidia is building global AI factories for this.
- Risks: Power constraints for giant data centers, dependence on TSMC in Taiwan, and inflated valuations of some companies are warning signs.
- **
