where is the money behind the AI boom actually going? a BNP Paribas breakdown shows how every $100 of illustrative AI infrastructure spending flows through the supply chain. > semiconductors take the largest share at $50. accelerators alone account for $25, followed by memory ICs at $15 and CPUs, other chips and server production at $10. these components provide the computing power needed to train and run AI models. > networking equipment takes $15. AI servers need high-speed connections to move data between chips and machines. switches, cabling and optical transceivers make that possible. > power accounts for $20. AI data centres need reliable electricity, from grid connections and independent generation to distribution equipment and on-site electrical systems. > cooling and facilities take another $15 combined. cooling accounts for $7.50, while construction and facilities account for another $7.50. both are essential for keeping large data centres operational. the bigger picture is that AI spending goes far beyond GPU makers. chip manufacturers capture a major share, but networking suppliers, power infrastructure and cooling companies are also benefiting from the expansion. as AI infrastructure grows, access to computing power, electricity and cooling could become just as important as the chips themselves. #BTC Price Analysis# #Bitcoin Price Prediction: What is Bitcoins next move?# #AI Agents 🤖# $BTC $ETH