Amazon Web Services (AWS) and OpenAI have formed a $38 billion partnership to enhance OpenAI's AI workloads. AWS will provide advanced computing resources, including NVIDIA GPUs and the ability to scale up to millions of CPUs, to support OpenAI's generative AI projects. The infrastructure is designed for high efficiency and low-latency performance.

SoftBank finalized its $40 billion investment in OpenAI, increasing its stake to about 11%. The funding includes a recent $22 billion tranche, aimed at supporting OpenAI's AI infrastructure and various projects, including a joint venture with Oracle. OpenAI is also preparing for an IPO and has attracted significant investment from Microsoft and Amazon.

Keith Heyde, newly appointed head of infrastructure at OpenAI, is leading the search for sites to build the company’s next-generation data centers, aimed at supporting the training of advanced AI models. With around 800 proposals received, about 20 sites are in advanced review, focusing on factors like power access and community support rather than just tax incentives. OpenAI's ambitious expansion includes a significant partnership with Nvidia, which is investing up to $100 billion to support the infrastructure needed for AI development.

OpenAI's CFO has indicated that the company is considering selling its infrastructure services to other firms, which could diversify its revenue streams beyond traditional product offerings. This move aligns with the growing demand for AI and machine learning capabilities among businesses.

OpenAI has struck a deal with AMD that allows the AI company to take a 10% stake in the chipmaker, while deploying 6 gigawatts of AMD's Instinct GPUs over the coming years. This partnership, which includes a warrant for up to 160 million shares, is set to alleviate OpenAI's compute power limitations and positions AMD as a key player in the AI industry. The deal also reflects the evolving interdependencies in the AI supply chain, with OpenAI actively building its infrastructure capabilities.