Podcast, Fireside Chat, Interview
How to Build a Thriving AI Ecosystem with Lisa Su, CEO of AMD
- AI is projected to represent the most significant technology discontinuity in the past 30 to 50 years, transforming society by enhancing productivity and intelligence, with widespread integration into all AMD products as a foundational element for future applications.
- Current generative AI adoption has accelerated but remains imperfect, requiring continued work to optimize technology across diverse form factors including edge, embedded, industrial, automotive, medical, client, and mobile environments.
- Supply chain constraints previously faced due to unforecasted generative AI needs, particularly in advanced packaging and high-bandwidth memory, highlight a need for better long-term demand forecasting, with new capacity requiring 18 to 24 months to deploy.
- Strategic focus is shifting toward building supply chain resilience rather than maximizing cost extraction, with the CHIPS Act prioritizing US manufacturing, R&D, and the training of future leaders to ensure a return on investment over the long term.
- Strategic planning relies on predicting disruptions three to five years out, with significant investments in advanced packaging, chiplets, and multi-node systems requiring sustained alignment with manufacturing partners like TSMC and early customer feedback cycles.
- AMD is pursuing an open ecosystem strategy to foster interoperability and standards like Ultra Accelerator Link, while utilizing M&A, such as the ZT Systems acquisition, to integrate hardware, software, and systems for high-performance computing clusters.
- The semiconductor landscape presents lower barriers to entry for startups due to access to world-class foundry capabilities, prompting AMD to actively collaborate with startups to leverage their pace of innovation and experimental agility.