Broadcom (AVGO) investor relations material

AI adoption and enterprise use cases

• 15,000 engineers engaged in exploring generative AI for hardware and software development, focusing on high-quality product creation and accelerated deployment.

• Key challenge is identifying use cases with significant ROI; productivity gains alone are often insufficient.

• AI tools are most effective for critical thinkers and senior architects, enhancing their capabilities.

• Early-stage adoption includes applications in product design and cybersecurity, with future potential in replicating software stacks.

Value creation in the AI stack

• Greatest long-term value expected to accrue to developers of leading frontier models due to exponential improvements.

• Model providers can capture significant revenue, with $20 billion accruing to model and application layers from a $30 billion ARR example.

Custom silicon and customer partnerships

• Custom accelerators (XPUs) are co-designed with frontier model developers for optimal performance and efficiency.

• Tight collaboration enables rapid development, exemplified by the Jalapeño chip for OpenAI delivered in nine months.

• Deep technical moats include advanced IP, high-speed interconnects, multi-die packaging, and out-engineering competitors.