GET IN EARLY! I'm Buying This AI Data Center Stock (Even Over NVIDIA Stock)

Here's a comprehensive summary of the provided YouTube video transcript:

Key Concepts

Coreweave (CRWV): A company providing high-performance computing infrastructure for the AI era.
AI Infrastructure: Data centers and computing resources specifically designed for AI workloads like training and inference.
Nvidia GPUs: Graphics Processing Units manufactured by Nvidia, essential for AI computations.
Revenue Backlog: The total value of committed future revenue from existing contracts.
Customer Concentration: The risk associated with a company relying heavily on a few large clients.
Capital Expenditures (CapEx): Investments made by a company in physical assets like data centers and equipment.
Stock-Based Compensation: Compensation provided to employees in the form of company stock.
Depreciation: The accounting process of allocating the cost of a tangible asset over its useful life.
Tensor RT LLM: An Nvidia software package that significantly improves the inference performance of large language models.
Picks and Shovels: A metaphor for companies that provide essential tools and infrastructure for a booming industry, rather than directly participating in the speculative aspect.

Coreweave's Business Model and Value Proposition

Coreweave positions itself as the "Amazon Web Services of the AI era," offering immediate access to advanced, high-performance computing infrastructure. This is crucial for companies that cannot afford the time or immense cost of building their own AI infrastructure.

Infrastructure: Coreweave operates 41 purpose-built AI data centers across North America and Europe, housing over 300,000 Nvidia GPUs. They possess access to 0.5 gigawatts of active power and 2.9 gigawatts of contracted power capacity. This contracted capacity is sufficient to power over 24,000 Nvidia Blackwell racks, equating to more than 1.7 million Nvidia GPUs.
Customer Base: Coreweave serves three primary customer segments:
- AI Labs: Developing foundational AI models, requiring massive parallel compute for training.
- Enterprises: Integrating AI into their applications and workflows, needing high-bandwidth, low-latency infrastructure for inference.
- Hyperscalers: Requiring overflow capacity during peak compute demand that exceeds their internal supply.
Competitive Advantage: Coreweave benefits from access to multiple Nvidia GPU generations (Hopper, Blackwell) and secures privileged access to Nvidia's latest chips 6-12 months before the general market.
Reliability as a Differentiator: Beyond raw capacity, Coreweave's key value lies in its reliability. It acts as a "peaker plant" for AI compute, providing the extra capacity needed to prevent "rolling compute blackouts" for AI workloads on major cloud platforms like AWS, Azure, and Google Cloud. This is particularly important because AI compute demand is less predictable and more volatile than traditional power demand, with fluctuations ranging from 30% for inference to over 90% for training.

Major Partnerships and Deal Values

Coreweave has secured significant multi-year contracts with major AI players:

OpenAI: Expanded their agreement by up to $6.5 billion this past quarter, bringing the total partnership value to over $22 billion. These deals ensure dedicated compute capacity for training and deploying advanced AI models, including reasoning models like GPT5. Reasoning models are noted for their high compute demands during inference due to complex problem-solving, code generation, and the "mixture of experts" approach.
Microsoft: Approximately $10 billion in total deal value to support Azure's AI services, including Bing AI and GitHub Copilot. Microsoft accounted for 62% of Coreweave's revenue in 2024 and an estimated 35% in 2025.
Meta Platforms: A $14.2 billion deal to provide dedicated compute clusters for training and running Llama models, and to scale Meta's overall AI infrastructure through 2031.
Nvidia: A formalized $6.3 billion partnership where Nvidia will purchase any unsold Coreweave cloud computing capacity through April 2032. This capacity will be used by Nvidia for their own AI labs, experiments, developers, and customers needing rapid access to large GPU clusters within Nvidia's CUDA ecosystem.

Latest Earnings and Financial Performance

Coreweave reported strong financial results, but also faced some headwinds that contributed to a stock price decline.

Record Revenues: $1.36 billion for the quarter, representing a 134% year-over-year increase.
Massive Revenue Backlog: Reached $55 billion, nearly quadrupling from the previous year. This backlog is driven by the large multi-year contracts with OpenAI, Microsoft, and Meta.
- 40% of the backlog is for 2-year contracts.
- 40% is for contracts between 2 and 4 years.
- 20% is for contracts of 4 years or longer.
Valuation Discrepancy: The speaker highlights that Coreweave's $55 billion backlog is 25% larger than its current market capitalization of $43 billion, suggesting significant undervaluation.
Capital Expenditures and Net Loss: The company spent $1.9 billion in capital expenditures in the past quarter, leading to a $110 million net loss.
Operating Margins: Collapsed from 20% last year to 4% this year.
Stock-Based Compensation: Increased significantly by 19x year-over-year, totaling $144 million, contributing to the net loss.

Reasons for Stock Price Crash and Risks

The stock price experienced a nearly 40% decline over two weeks, attributed to several factors:

Decreased 2025 Revenue Guidance: Coreweave reduced its 2025 revenue guidance by 3%. This revision, implying around 9% lower expectations for Q4, was due to construction delays at a single data center out of 41. The issue is expected to be resolved by Q1.
Michael Bur's AI Bubble Short: Prominent investor Michael Bur has bet over a billion dollars that AI is the next bubble to pop. His argument centers on the alleged underreporting of depreciation by AI infrastructure companies, who may be overstating the useful lifespan of AI chips to boost earnings. He estimates this could understate depreciation by $176 billion across the industry, potentially overstating profits for companies like Meta and Oracle.
- Counter-Argument on Depreciation: The speaker counters Bur's argument by referencing Nvidia's Tensor RT LLM software. This software doubled inference performance for large language models on GPUs, effectively extending the useful lifespan of older chips (like Hopper H100s and Ampere A100s) and current Blackwell GPUs. The speaker argues that if software updates can extend a chip's operational life, it supports longer depreciation periods, challenging Bur's claims of accounting fraud.
High Capex and Stock-Based Compensation: The significant capital expenditures and stock-based compensation leading to a net loss for the quarter are also cited as contributing factors.

Investment Thesis and Conclusion

Despite the recent stock price drop and short-term risks, the speaker presents a strong bullish case for Coreweave.

Nvidia's Largest Public Position: The fact that Nvidia holds Coreweave as its largest public position suggests Nvidia recognizes the long-term value of guaranteed, reliable AI infrastructure in a supply-constrained market.
Predictable Future Revenues: The $55 billion backlog provides exceptional visibility into future growth.
Growing Backlog vs. Spending: The backlog is growing at a faster rate than the company's spending.
Strategic Spending: The capital expenditures are directly invested in business growth, including acquiring/building data centers, securing power, and expanding the workforce.
"Picks and Shovels" Analogy: Coreweave is described as a company selling "picks and shovels" during an AI gold rush, a historically profitable business model.
Undervaluation: The stock is trading significantly below its calculated fair value, with a discounted cash flow model suggesting it is 80% undervalued and would need to increase by 400% to reach fair value.

The speaker concludes that the current stock price decline presents a significant opportunity for long-term investors, driven by predictable revenue, strategic growth investments, and a strong market position.