We didn't see this coming. But the market's reaction to Core Scientific's rejection of a $9 billion buyout is a textbook case of narrative over substance. The stock jumped 12% on the AMD partnership announcement, and the media spun it as a vote of confidence in the company's AI pivot. Let me be clear: rejecting a $9B exit is not a sign of strength. It's a bet that the company can create more value from scratch – and that bet hinges on a technology partnership that has delivered exactly zero proof points.
Context: Core Scientific, one of the largest Bitcoin miners, emerged from Chapter 11 in early 2024 with a restructured balance sheet and a new strategy – pivot from mining to AI and high-performance computing (HPC) hosting. The company signed a multi-year hosting deal with CoreWeave in 2024, then announced a partnership with Advanced Micro Devices (AMD) in early 2025. The AMD partnership was supposed to accelerate the AI pivot. Then, a $9 billion acquisition offer came in – reportedly from a private equity consortium – and shareholders rejected it. The official reason: the board believes the AMD partnership and AI pivot will generate more long-term value. The market cheered. But the technical details tell a different story.
Let's dig into the core: The AMD partnership is a press release, not a deliverable. The announcement contained no technical specifications, no committed megawatts, no GPU delivery timelines, and no revenue guarantees. Based on my experience auditing data center infrastructure for a major exchange, I can tell you that transitioning from ASIC mining to GPU AI workloads is like replacing a car engine mid-flight. It's not plug-and-play. Bitcoin mining rigs are simple – they consume power, hash, and produce BTC. AI data centers require liquid cooling, high-density racking, InfiniBand or RoCE networking, and a software stack that can orchestrate distributed training. Core Scientific has none of that proven at scale. Their mining facilities are designed for low-density, high-power ASICs, not for the thermal and networking demands of AMD Instinct GPUs.
s evolution of the company from mining to AI is not a linear path; it's a capital-intensive transformation that could destroy shareholder value if execution fails. The market is pricing the option value of the AI pivot, not the underlying risk. Consider the capital structure: Core Scientific still carries debt from its bankruptcy restructuring, with covenants that likely restrict capital spending. To retrofit facilities for AI, the company will need to raise additional capital – either through equity dilution or convertible debt. The $9B rejection sets a high bar: the company must now create $9B+ in equity value from a business that is still majority Bitcoin mining (with volatile revenue) and an AI hosting business that is unproven. The AMD partnership does not guarantee revenue; it's a vendor relationship, not a revenue-sharing contract. The real value in Core Scientific is its power contracts – cheap, long-term electricity locked in for mining. But AI data centers require constant, high-utilization power, not the interruptible load that mining uses. The grid interconnection may not support the shift.
We didn't see any mention of liquid cooling, high-density racking, or software stack. The AMD partnership is a hardware supply agreement, but AMD's Instinct GPUs still lag behind Nvidia's CUDA ecosystem in software maturity. ROCm is improving, but it's not production-ready for the largest AI workloads. Core Scientific becomes a beta tester for AMD. If the GPUs fail, Core Scientific's AI dreams evaporate. If they succeed, AMD captures the value through increased chip sales. Shareholders are betting on a technology that is not yet proven.
Contrarian: The unreported angle is that this deal is a win for AMD, not for Core Scientific. AMD needs real-world deployment sites to validate its Instinct GPUs against Nvidia's dominance. Core Scientific provides that laboratory – at its own expense. The market is ignoring the asymmetrical risk: AMD gets a free testing ground, while Core Scientific bears the capital cost and operational risk. The media is also missing the capital structure trap. Rejecting $9B means the company must now deliver $9B+ in value creation. But the path to that value is through a high-risk technical pivot with no guarantees. The market is pricing the narrative, not the numbers. We didn't see any discussion of the joint development agreement terms – who bears the cost if the deployment fails? That's the key.
Takeaway: Watch the next quarterly filings for two numbers: AI revenue (if any) and capital expenditure. If they raise capital or issue convertible notes, it's a signal that the AMD partnership is not self-funding. The question is not whether Core Scientific can pivot to AI – it's whether the market will realize that this pivot is a multi-year gamble before the first GPU is even racked. When the next quarterly report shows zero AI revenue, will the market still believe the narrative?
