AMD and Core Scientific have today announced an AI infrastructure partnership to support growing demand for AMD compute.
The alliance is designed to expand the AMD ecosystem, with the chip giant expected to secure up to 2.5GW of data centre capacity to support end customer deployments of its solutions. Also part of the agreement, both companies will partner on physical infrastructure design and deployment of AMD Instinct GPUs, EPYC CPUs and ROCm software.
It also supports Core Scientific’s shift from crypto towards hosting AI and high-performance computing (HPC) workloads.
“We are proud to establish a strategic relationship with AMD and support the continued deployment of its next-generation products,” said Adam Sullivan, CEO of Core Scientific. “Our proven execution capabilities and ability to deliver high-density infrastructure at scale position us to support AMD’s technology roadmap and grow our relationship meaningfully over time.”
The agreement means that AMD has successfully secured more than 500MW of US capacity beginning in 2027. By expanding its access to Core Scientific’s AI-ready infrastructure, AMD expects to enable its end customers to deploy AI faster and at a greater scale.
This comes as AI demand only continues to accelerate, making access to land, power and data centre infrastructure critical to fuelling the next chapter of AI compute. In AMD’s mind, bringing that compute online requires strong partnerships with infrastructure experts who can support the scale and power needed to enable the next phase of AI development.
“Core Scientific’s extensive portfolio of AI-ready data centres expands access to the infrastructure our customers need to deploy AMD AI solutions at scale,” shared Mathew Hein, senior vice president and chief strategy officer of corporate development at AMD. “Together, we are helping model builders, cloud providers and enterprises accelerate AI adoption while strengthening the AMD ecosystem.”
As part of the agreement, AMD will also receive market-priced warrants to purchase Core Scientific’s common stock, subject to certain commercial conditions.
AMD has announced a series of AI-intensive deals in recent days, which coincided with its Advancing AI event in San Francisco last week. It released its first jointly developed reference design for the AMD Helios rackscale platform with Schneider Electric, designed to speed up deployment of high-density AI infrastructure. Supermicro has also introduced a new H15 server portfolio with AMD CPUs to deliver stronger performance improvements.
The company was also reported to consider signing a multi-year CPU purchase commitment with Intel, according to Reuters, to lock in as AI demand drives up prices and increases shortages.
Related stories
AMD powers 41% of new Top500 systems and more than half of Green500’s top 50
Meta and BlackRock pen AI data centre agreement in Texas
Microsoft to deploy AMD’s next-gen Helios platform in Azure AI infrastructure expansion






