I'm Buying THIS AI Stock To Get Rich (Without Getting Lucky)

Key Concepts

• Coreweave (CRWV): An AI infrastructure company providing GPU compute power.
• Nvidia: A major investor in Coreweave, holding a significant portion of its stock.
• AI Infrastructure: The hardware and services required to train and run artificial intelligence models.
• GPU (Graphics Processing Unit): Specialized processors crucial for AI computations.
• Hyperscalers: Large cloud computing providers like AWS, Azure, and Google Cloud.
• AI Labs: Companies focused on developing foundational AI models.
• Enterprises: Businesses implementing AI applications.
• Model Training: The process of teaching AI models using large datasets.
• Inference: The process of using a trained AI model to make predictions or generate outputs.
• Reasoning Models: Advanced AI models that can break down problems and solve them step-by-step.
• Stock-Based Compensation: Paying employees with company stock.
• Capex (Capital Expenditures): Money spent on acquiring or upgrading physical assets.
• Adjusted EBITDA: Earnings Before Interest, Taxes, Depreciation, and Amortization, adjusted for certain items.
• IPO Lock-up Period: A period after an IPO during which early investors cannot sell their shares.
• Backlog: The total value of signed multi-year contracts for future services.
• Peaker Plants: Power plants that operate only during periods of high demand.
• Liquid Cooling: Advanced cooling systems for high-density computing hardware.
• Discounted Cash Flow (DCF) Model: A valuation method that estimates the value of an investment based on its expected future cash flows.

Coreweave: The AI Infrastructure Powerhouse

This analysis delves into Coreweave (ticker symbol CRWV), an AI infrastructure company that has attracted significant attention, including a substantial investment from Nvidia. The video aims to explain Coreweave's business model, recent financial performance, associated risks, and its current valuation.

Coreweave's Business Model: The AWS of AI

Coreweave positions itself as the "Amazon Web Services of AI," providing immediate access to advanced GPU infrastructure for companies that cannot afford or wait to build their own.

• Infrastructure: Coreweave operates 33 data centers across North America and Europe, housing over 4 million Nvidia GPUs. They possess access to nearly 0.5 gigawatts (GW) of active power and have contracted for an additional 2.2 GW.
• Customer Segments:
  • AI Labs: Require massive compute for foundational model training.
  • Hyperscalers: Need overflow compute capacity to supplement their existing infrastructure.
  • Enterprises: Seek straightforward solutions for implementing AI applications, particularly for inference.
• GPU Access: Coreweave utilizes multiple Nvidia GPU generations, including Hopper and Blackwell, and often secures privileged access to new architectures 6-12 months before general market release, providing a significant competitive edge.
• Key Partnerships:
  • OpenAI: Has a $12 billion, 5-year agreement, expanded by an additional $4 billion. This provides dedicated compute for training and deploying models like GPT-5, which are computationally intensive due to their reasoning capabilities. Reasoning models utilize more compute during inference by breaking down problems, writing code, or employing a "mixture of experts" approach.
  • Microsoft: Coreweave's largest customer, accounting for over half of their 2024 revenue. Microsoft increased its spending by 8x from a substantial $230 million deal in 2023. Microsoft utilizes Coreweave's infrastructure when Azure's internal capacity is exceeded for services like GitHub Copilot, Bing AI, and enterprise AI applications.
• Addressing Peak Demand: AI data centers experience significant power demand fluctuations (30% of peak for inference, 90%+ for training). Coreweave acts like "peaker power plants" for compute, providing the extra capacity needed during peak AI training workloads to prevent "compute blackouts."

Acquisition of Core Scientific and Market Reaction

Coreweave's recent acquisition of Core Scientific, a former bankrupt Bitcoin miner now repurposed for AI infrastructure, is highlighted.

• Core Scientific's Assets: The acquisition brings over 1.3 GW of contracted power capacity across nine US data centers and over $10 billion in secured contracts for AI and high-performance computing. Core Scientific also boasts industry-leading liquid cooling systems and direct utility partnerships for low-cost power.
• Synergy and Backlog: The acquisition is seen as highly synergistic, helping Coreweave address its $30 billion backlog of demand. This backlog represents signed, multi-year commitments for their AI infrastructure and data center services, exceeding half of Coreweave's current market capitalization.
• Market Misinterpretation: The video argues that the market's negative reaction (stock crash) to the Core Scientific acquisition is a mistake, as it overlooks the significant synergies and the immense demand backlog.

Coreweave's Financials and Valuation

The video examines Coreweave's recent earnings and its perceived undervaluation.

• Revenue Growth: Coreweave reported $1.2 billion in revenue for the quarter, a 207% increase year-over-year. Revenue has consistently grown by 30-50% quarter-over-quarter.
• Backlog Growth: The revenue backlog reached $30 billion, an 86% year-over-year increase, driven by contracts with OpenAI and a hyperscaler (likely Microsoft).
  • 50% of the backlog is for 2-year contracts.
  • 40% is for 2-4 year contracts.
  • 10% is for 4-6 year contracts.
• Risks and Stock Performance:
  • High Capex: $2.9 billion in capex this quarter, nearly 2.5 times revenue, led to a $290 million net loss.
  • Stock-Based Compensation: A 20x year-over-year increase in stock-based compensation, contributing to a collapse in operating margins from 20% to 2%.
  • Selling Pressure: The end of the IPO lock-up period on August 15th allowed early investors to sell shares that had appreciated significantly (100-200% since IPO), creating selling pressure.
• Mitigating Factors:
  • Capex Investment: The high capex is viewed as necessary investment in scaling the platform to meet the $30 billion demand backlog, ensuring future returns.
  • Acquisition Rationale: The Core Scientific acquisition is seen as strategic for securing power and data center capacity.
  • Adjusted EBITDA: $750 million, tripling year-over-year, with a 62% adjusted EBITDA margin, indicating strong unit economics once infrastructure is deployed.
  • Stock-Based Compensation: Considered standard for fast-growing AI companies to attract and retain talent in a competitive market.
  • Insider Selling: Typical post-IPO behavior as early investors realize gains.
• Undervaluation: Discounted cash flow models suggest a fair value of around $220 per share, implying Coreweave is 55% undervalued at its current price, with potential for 120% upside.

Conclusion: A Strategic Investment Opportunity

Coreweave is presented as a "picks and shovel" supplier in the AI gold rush, providing essential infrastructure. The video argues that the recent stock crash is an overreaction to short-term financial metrics and selling pressure, overlooking the company's strong fundamentals, massive backlog, and strategic investments. The predictable revenue model based on multi-year contracts, coupled with significant growth in demand, positions Coreweave as a compelling investment opportunity for those seeking to "get rich without getting lucky."