Citigroup recently upgraded AMD's stock to 'Buy' and significantly raised its price target, signaling strong confidence in the company's future in the AI era.
This optimism is built on two powerful narratives. The first is AMD's solid establishment as a 'second source' for AI GPUs. For a long time, Nvidia has dominated this market, but major tech companies, or hyperscalers, are eager for a reliable alternative to mitigate supply risks. AMD solidified its position with a massive, multi-year agreement to supply up to 6 gigawatts of its Instinct GPUs to Meta. This deal, potentially worth tens of billions of dollars, provides unprecedented visibility into AMD's future data center revenue, which is a key reason for Citi's bullish stance.
But the story isn't just about GPUs. The second narrative is the dawn of a 'CPU renaissance,' driven by the rise of 'agentic AI.' Unlike current AI that typically handles single tasks, agentic AI performs complex, multi-step workflows. This requires significantly more processing power from server CPUs, not just GPUs. This trend is expected to dramatically expand the market for server CPUs, turning AMD's EPYC processor lineup into a second, powerful growth engine independent of its GPU success.
This upgrade wasn't a sudden event; it was the culmination of a series of strategic moves. It began with foundational deals with OpenAI and Meta, which established AMD's credibility. Then, the ecosystem began to form, with partners like Supermicro building full server systems around AMD's technology, proving they could deliver at scale. Finally, the ongoing global shortage of HBM (High Bandwidth Memory), a critical component for AI chips, has made hyperscalers even more motivated to secure a stable supply from a second source like AMD.
However, challenges remain. AMD's success heavily depends on its ability to execute, particularly in securing a sufficient supply of HBM and ensuring its software platform, ROCm, is mature and competitive. All eyes are now on AMD's upcoming 'Advancing AI 2026' event in July, which could provide further proof of customer adoption and technological progress.
- Glossary
- Agentic AI: An advanced form of AI capable of autonomously performing complex, multi-step tasks and making decisions to achieve a specific goal.
- Hyperscaler: A large-scale cloud service provider, such as Amazon Web Services (AWS), Microsoft Azure, Google Cloud, and Meta, that operates massive data centers.
- HBM (High Bandwidth Memory): A type of high-performance memory used in GPUs and other AI accelerators, essential for handling the large datasets required by AI models.