AMD guarantees $300M Crusoe loan to accelerate AI chip deployment
Chipmaker mirrors Nvidia financing strategy with lease-back clause to secure chip demand
Advanced Micro Devices has agreed to guarantee a $300 million loan for AI infrastructure startup Crusoe, arranged by Goldman Sachs, in a strategic move that mirrors Nvidia's playbook of financing customers to drive chip adoption. AMD shares rose 1.6% to $203.37 in Thursday trading as investors welcomed the aggressive move to gain ground in the fiercely competitive artificial intelligence chip market.
The loan will enable Crusoe, which pivoted from cryptocurrency to AI infrastructure after its founding in 2018, to purchase AMD's AI accelerators for deployment in a new Ohio data center. The data center is being developed by Canadian firm 5C with backing from Brookfield, a major infrastructure investor. AMD's guarantee helped Crusoe secure an interest rate of approximately 6%, significantly lower than what the startup would have obtained in the open market.
Crucially, the agreement includes a lease-back provision that allows AMD to repurchase the chips if Crusoe fails to attract sufficient customers from AI developers, limiting AMD's risk exposure while ensuring its hardware finds a home. The loan is collateralized by the chips and associated equipment, providing additional security for Goldman Sachs. Crusoe plans to purchase a total of $400 million worth of AMD chips, indicating a broader commitment beyond the initial loan facility.
The strategy marks a significant escalation in AMD's competitive posture against Nvidia, which maintains an estimated 80-85% market share in the AI data center segment. Nvidia has employed similar financing arrangements with AI cloud providers like CoreWeave to accelerate adoption of its graphics processing units. AMD's move comes as it seeks to capitalize on growing demand for alternatives to Nvidia's dominant position in the AI accelerator market.
AMD's hardware has become increasingly competitive, with its MI350 and MI355X series entering volume production in June 2025 and positioning as challengers to Nvidia's Blackwell architecture. The company claims its MI355X delivers up to 40% better "tokens-per-dollar" performance than Nvidia's GB200, offering a compelling value proposition for cost-conscious AI developers. Looking ahead, AMD's MI450 GPUs are slated for deployment with OpenAI in the second half of 2026, signaling growing confidence from major AI players.
However, analysts caution that AMD still faces challenges in matching Nvidia's full-stack advantage. While AMD's hardware specifications have narrowed the gap, some analysts highlight AMD's software ecosystem as a current weakness compared to Nvidia's mature CUDA platform and comprehensive networking infrastructure. This software gap can impact real-world performance and make switching costs higher for potential customers.
The Crusoe deal represents AMD's recognition that financial engineering is becoming as important as technological innovation in the AI chip wars. By providing financing support, AMD reduces the capital burden on customers who might otherwise struggle to afford upfront purchases of expensive AI accelerators. This approach can accelerate adoption rates and help AMD secure critical design wins with AI companies that could become long-term partners.
From a financial perspective, the loan guarantee carries limited downside risk for AMD given the lease-back clause and equipment collateral. The primary risk is opportunity cost—the chips deployed with Crusoe could potentially be sold elsewhere at higher prices. However, securing deployment in a new data center provides valuable reference customers and real-world performance data that AMD can leverage to attract additional business.
AMD's fundamentals support its ambitious AI push. The company reported 34.1% year-over-year revenue growth and 217.1% earnings growth in its most recent quarter, with analysts maintaining an average price target of $287.52—representing more than 40% upside from current levels. The stock trades at a forward price-to-earnings ratio of 19.09, suggesting investors are already pricing in significant growth expectations from the AI business.
The Crusoe transaction comes at a critical juncture in the AI infrastructure build-out. Major cloud providers and AI companies are racing to secure computing capacity to train increasingly large language models. By aligning with Crusoe, which is building greenfield data center capacity, AMD positions itself to benefit from the next wave of AI infrastructure spending rather than competing solely for share in existing facilities dominated by Nvidia hardware.
Looking ahead, investors will be watching whether AMD expands this financing approach to additional customers and whether the Crusoe deployment succeeds in attracting AI developers. Success could validate a broader strategy of using financial levers to accelerate market share gains, while failure could highlight the challenges of displacing Nvidia's entrenched ecosystem advantage.