Vote to see vote counts
OpenAI and Broadcom have announced a multi-billion dollar deal to develop 10 gigawatts of custom AI chips over the next four years. This collaboration will enable OpenAI to design its own GPUs, powering both its data centers and those of its partners. The chips are expected to roll out in the second half of next year.
OpenAI's partnership with NVIDIA involves a $100 billion investment aimed at expanding compute capacity.
NVIDIA's partnership with OpenAI involves investing $100 billion over time, potentially generating $400 billion in revenue for NVIDIA.
NVIDIA's $100 billion investment in OpenAI will allow OpenAI to use NVIDIA's AI semiconductors, requiring 10 gigawatts of power, equivalent to the consumption of 8 million homes.
The partnership between NVIDIA and OpenAI involves a $100 billion investment to build AI infrastructure, potentially generating $400 billion in revenue for NVIDIA.
OpenAI has struck a multi-billion dollar deal with AMD, securing 6 gigawatts of compute capacity to power its AI data centers. In exchange, OpenAI could secure up to a 10% stake in AMD if it meets certain deployment milestones.
OpenAI's compute commitments now exceed $1.2 trillion, which is five times what all of big tech plans to spend on CapEx this year. This is also more than 90 times larger than OpenAI's projected $13 billion in revenue for this year.
The AI money bubble is a significant concern, with over a trillion dollars of incremental CapEx being announced, such as the recent OpenAI and Broadcom deal. This level of capital expenditure raises red flags about how it's being financed.
OpenAI's cash burn is estimated at $115 billion through 2029. That's an unbelievable amount of money!
The recent OpenAI-Broadcom deal is valued at approximately $350 billion, or $88 billion per year, based on the estimation that one gigawatt of chips equals about $35 billion. This deal adds to OpenAI's existing commitments with AMD, Nvidia, CoreWeave, and Oracle, totaling around $880 billion.