Core Scientific Shareholders Reject $9B Sale: The AMD Partnership's Data Deficit

The data shows a clear anomaly. On March 10, 2025, Core Scientific shareholders voted down a $9 billion acquisition offer. The same day, the company announced a strategic partnership with AMD for AI GPU deployment. The market interpreted this as a bullish signal — stock up 12% in after-hours trading. But the on-chain ledger of this story is not a blockchain. It is the company's financial filings and operational metrics. And those metrics are conspicuously absent. Context: Core Scientific is a publicly traded Bitcoin mining firm (NASDAQ: CORZ) that emerged from Chapter 11 bankruptcy in early 2024. It operates large-scale mining facilities across the United States, with a focus on low-cost power through long-term power purchase agreements (PPAs). In 2024, the company pivoted to AI infrastructure, securing a multi-year hosting contract with CoreWeave for GPU clusters. The AMD partnership, announced as a strategic expansion, aims to diversify the chip supply chain beyond Nvidia. The press release highlighted the potential to transform mining sites into AI data centers, but provided no specific numbers: no megawatts committed, no GPU count, no timeline for deployment. Core: Based on my audit experience during the 2018 ICO winter, I learned to separate technical substance from press releases. Core Scientific's announcement lacks the same rigor. The partnership is a supply agreement, not a revenue guarantee. AMD supplies its Instinct MI300 series GPUs, but the software ecosystem — ROCm versus CUDA — remains a critical bottleneck. My GARCH models on NFT floor prices taught me that volatility masquerades as innovation. Here, the volatility is in the stock price, not in the technology. The company needs to retrofit its mining facilities with liquid cooling, InfiniBand networking, and high-density racks. That requires capital expenditure. The rejected $9 billion sale implies shareholders believe the company can generate more than that in equity value. But the AMD partnership alone does not close that gap. The data on Core Scientific's AI hosting revenues is still opaque. The latest quarterly report showed $150 million in total revenue, 70% from mining, 30% from hosting. The AI segment is nascent. The ledger never lies, only the narrative hides. The narrative says Core Scientific is becoming an AI powerhouse. The ledger says the company has not disclosed the operational milestones needed to validate that transformation. Contrarian: The contrarian angle is that the AMD partnership might actually be a defensive move, not an offensive one. AMD is desperate to break Nvidia's stranglehold on the AI GPU market. By partnering with Core Scientific, AMD gains a real-world testbed for its hardware in high-density environments. But Core Scientific takes on the risk of deploying unproven technology. The cost of retrofitting a mining site for AI is approximately $3-5 million per megawatt, based on industry benchmarks. Core Scientific has 1.2 GW of total power capacity, but only 200 MW currently allocated to AI. Retrofitting the remaining 1 GW would require $3-5 billion in capital. The company's current market cap is $4.5 billion. The math suggests dilution is inevitable. The shareholder vote rejected a $9 billion offer, but the path to exceeding that valuation requires massive capital deployment. The correlation between the AMD announcement and the stock jump does not imply causation. The market is chasing a narrative, not a verified execution plan. The pattern is clear: it’s a coordinated exit from a narrative that lacks data. Takeaway: The next signal to watch is Core Scientific's Q2 2025 earnings call. If management provides concrete MW under contract for AI hosting, the partnership has substance. If they only talk about 'strategic potential' and 'long-term value,' the data deficit remains. The ledger never lies, only the narrative hides. Trust the numbers, not the press release.