Core Scientific’s $9B Rejection: The AMD Partnership That Isn’t There

The shareholders of Core Scientific did something unusual in a market starved for conviction. They said no to nine billion dollars. The offer was real. The acquirer was credible. The premium was substantial. Yet the board and a majority of holders rejected the deal, pointing instead to a new partnership with AMD as the foundation for greater value. The stock barely moved. The market is waiting for proof. s heart. Here is the context: Core Scientific is a Bitcoin mining company that survived Chapter 11 bankruptcy in 2023 by pivoting to AI data center hosting. The pivot is real. They signed a multi-year contract with CoreWeave, a pure-play AI cloud provider, to lease out their power capacity. They have existing infrastructure—sites with high-voltage power lines, cooling systems, and security. The mining business still runs, but the AI hosting represents the growth narrative. The AMD partnership, announced in late 2024, was positioned as the next step: a strategic alliance to deploy AMD Instinct GPUs in their facilities, diversifying away from Nvidia-dependent CoreWeave deals. The press release was thin. No capacity numbers. No timeline. No revenue commitment. The market reacted with a shrug. The shareholders' rejection of the $9B acquisition set a valuation anchor: the board believes the company is worth more than that. But the AMD partnership is the only visible catalyst. And it is a vessel of air. Let me dissect the technical reality. In my eight years of auditing crypto infrastructure projects, I have learned to distinguish between a press release and a product. The AMD partnership is a press release. The core of the announcement is a statement of intent to "collaborate" on deploying AMD GPUs. There is no mention of how many GPUs, which specific models, or when they will be operational. There is no reference to any existing test deployment, benchmark results, or customer commitments. The only concrete detail is that AMD will supply chips. That is the minimum bar for a partnership announcement. It is not a milestone. It is a starting line that has not been crossed. Let me compare this to the only other comparable pivot in the space: Hive Blockchain’s transformation into Hive Digital Technologies. Hive, like Core Scientific, was a mining firm that shifted to AI hosting. But Hive provided quarterly updates on GPU count, utilization rates, and revenue per GPU. They published their power costs and the types of customers they were serving. Core Scientific, in contrast, offers only narrative. The AMD partnership is a blank check on hype. The company has not disclosed the total power capacity dedicated to AI, the number of deployed GPUs (from any vendor), or the utilization of those GPUs. These are the metrics that matter. Without them, the announcement is a marketing artifact. The engineering challenge here is significant. Converting a Bitcoin mining site to an AI data center is not a matter of swapping ASICs for GPUs. Mining sites are designed for high-density, low-latency power draw but with minimal networking requirements. AI workloads, especially training, require InfiniBand or RoCE networking, liquid cooling for high-performance GPUs, and a different physical layout to accommodate server racks with higher heat density. Core Scientific’s existing sites may have the power capacity, but they do not have the network infrastructure. The cost of retrofitting is non-trivial. The timeline for such retrofits is measured in months, not weeks. The AMD partnership provides no details on how these engineering challenges will be addressed. The press release says "collaborate on engineering optimization"—a phrase that means nothing without a timeline or budget. Then there is the software ecosystem. AMD’s ROCm stack has improved significantly, but it still lags CUDA in terms of developer tooling, library support, and performance optimization for many AI models. A partnership with AMD does not automatically make the GPUs usable. Core Scientific will need to invest in software engineering to ensure that their customers’ workloads can run efficiently on AMD hardware. This is a cost center, not a revenue stream. The announcement does not mention any software partnership or investment. It is a hardware supply deal, and hardware alone does not produce revenue. Now consider the financial structure. Core Scientific is a public company with a debt burden from its bankruptcy restructuring. The $9B acquisition offer was likely a cash deal that would have wiped out equity holders or given them a premium. The rejection implies that management believes the AI pivot can generate more than $9B in net present value. That is a bold bet. The company’s current market cap is around $5-6B, so the offer was a significant premium. To justify the rejection, the board must deliver on the AMD partnership. But the partnership has no revenue guarantees. It is a procurement agreement, not a revenue contract. The value of the partnership will be determined by the actual sales of AI hosting services using AMD hardware. Those sales depend on demand, which is currently concentrated on Nvidia. The market is already seeing a shift toward AMD, but it is gradual. Core Scientific is betting on a wave that is not yet visible. The tokenomics frame does not apply here—CORZ is a stock, not a token. But the same principles of value capture apply. There is no built-in deflation mechanism. No staking yields. No fee burning. The only way for shareholders to profit is through operational cash flow and multiple expansion. The AMD partnership is a narrative lever for multiple expansion, but it is not yet backed by cash flow. The $9B rejection sets a floor on the narrative: the company must be worth more than that. But the market will test that floor with every quarterly report. If the next earnings report shows no increase in AI hosting revenue or capacity, the stock will fall. The AMD partnership will be exposed as a theatrical prop. I have seen this pattern before. In 2021, I audited the contracts of a mid-tier NFT project that claimed to use IPFS for metadata storage. The marketing said "fully decentralized." The reality was that 70% of the assets were stored on a centralized server that could be taken down with a single DMCA request. The gap between the claim and the proof was the same as the gap between Core Scientific’s AMD partnership announcement and any technical validation. The market is currently trading on the claim, but the proof is missing. s heart. Empty metadata, full wallets. Let me address the contrarian angle. The bulls might argue that Core Scientific’s power infrastructure is a real asset that cannot be replicated. They have long-term power purchase agreements at fixed rates, which gives them a cost advantage over cloud providers that buy power on the spot market. AI workloads are energy-intensive, and power cost is a major component of total cost of ownership. Core Scientific could become a low-cost provider of AI compute. The AMD partnership could diversify their supply chain, reducing dependency on Nvidia, which is increasingly competitive. Additionally, the $9B rejection could signal that management sees a clear path to a higher valuation, perhaps through a future sale or a larger deal. The bulls might also point to the growing demand for AI inference, which is less sensitive to GPU ecosystem lock-in than training. Core Scientific could focus on inference workloads, which are more tolerant of AMD’s software stack. These arguments have merit. The power infrastructure is real. The cost advantage is real. The demand for AI compute is real. But the contrarian view must be weighed against the lack of execution evidence. The AMD partnership is the only concrete step toward capturing that value, and it is a step that has not yet been taken. The bulls are betting on the potential, not the reality. The market is currently pricing in a discount for that uncertainty. The question is whether the discount will shrink as the company delivers on its promises. The answer depends on the next 12 months. Takeaway: The Core Scientific board has made a high-stakes bet. They are asking the market to trust that the AMD partnership will transform the company. The evidence so far is a press release with no technical detail. The next earnings report will be the first test. If the company discloses any meaningful operational metrics—MW of AI capacity, GPU count, utilization rate, or revenue from AI hosting—the narrative will strengthen. If not, the rejection of the $9B offer will look like a mistake. The smart money is watching the data, not the headlines. s heart. The announcement is a narrative hedge, not a transformation. The real story is still unwritten.