• Three-pillar AI strategy: AMD is driving its AI roadmap by delivering leadership hardware on an annual product cadence, collaborating closely with the open-source ecosystem, and partnering with planetary-scale leaders to co-design future roadmaps.
• Inference cost curve: Hardware efficiency is improving by roughly 10x generation-over-generation, which will rapidly lower per-token costs as the industry shifts its primary monetization focus from model training to pervasive inference.
• Agentic compute demand: The rise of autonomous agentic workflows is increasing demand for CPUs to actively coordinate and orchestrate high-speed, agent-to-agent task execution alongside GPUs.
• India engineering footprint: AMD employs over 10,000 people in India who contribute design and software across the entire product portfolio, while also seeing growing opportunities in the country's expanding data center ecosystem.