On November 22, foreign media reported that Amin Vahdat, the leader of Google Cloud's AI infrastructure division, revealed at a company-wide meeting that to keep pace with the skyrocketing demand for AI services, Google must double its infrastructure service capacity every six months. The company has set an ambitious goal of achieving a 1,000-fold increase in capacity within the next four to five years. Google has recently unveiled its seventh-generation TPU chip, codenamed 'Ironwood,' which boasts an energy efficiency 30 times greater than that of its first-generation counterpart. Additionally, the company has harnessed the power of DeepMind to enhance model efficiency. Alphabet, Google's parent company, has raised its capital expenditure forecast for the second time this year, now projecting expenditures between $91 billion and $93 billion, with a notable surge anticipated in 2026. Google CEO Sundar Pichai acknowledged market concerns regarding a potential AI bubble but emphasized that the risk of underinvestment outweighs such worries. He justified the substantial investments by pointing to the revenue growth observed in cloud services.