- Microsoft and OpenAI consider a $100 billion data center for AI development.
- “Stargate” project could become the most expensive, aiming for launch by 2028.
- Project includes multi-phase plans with significant investments in AI server chips.
In a bold collaboration, Microsoft and OpenAI sketch out a visionary project dubbed “Stargate,” aspiring to erect a supercomputer for unparalleled AI innovation. With an eye-watering budget of up to $100 billion, this initiative underscores the duo’s commitment to spearheading AI’s frontier, earmarked by the ambition to dramatically expand computing capabilities for AI research and development. The envisioned data center, potentially based in the U.S., is to be outfitted with millions of server chips, signifying a leap in computational power necessary for advancing artificial general intelligence.
The groundwork for “Stargate” reflects Microsoft’s strategic investment in OpenAI, surpassing $13 billion to date, facilitating OpenAI’s endeavors like ChatGPT and ensuring Microsoft’s exclusive access to OpenAI’s cutting-edge technologies. The collaboration anticipates the supercomputer’s culmination by 2030, requiring gigawatts of power and a robust chip inventory, with considerations for energy sourcing posing a critical challenge.
This unprecedented venture into AI infrastructure reveals a phased strategy, with “Stargate” as the pinnacle phase five, following a series of progressively ambitious supercomputing projects. Current dialogues hint at a potential pivot towards alternative chip manufacturers to mitigate reliance on Nvidia, driven by escalating demand and costs. Amidst technical and logistical complexities, including enhanced GPU density and network architecture considerations, “Stargate” promises a quantum leap in AI’s computational landscape, aligning with visions of tackling humanity’s most daunting challenges through superintelligence.