According to a research report released by Barclays Bank, the AI sector is still far from entering a phase characterized by a bursting bubble and a slowdown in investment. The capital expenditure cycle within the AI industry is set to persist, with pivotal technological advancements anticipated to spark a substantial surge in demand for computing power between 2027 and 2028. The fundamental premise for this forecast lies in OpenAI's impressive revenue performance, which has outstripped expectations. Its revenue for 2025 has already surpassed the internal projections made in mid-year, and the projected revenue for 2027 has been revised upwards, indicating the potential to achieve the annual recurring revenue target of USD 100 billion a year earlier than planned. This accomplishment has not only enhanced key financial metrics but also mitigated the risk of an industry-wide bubble burst. OpenAI's voracious appetite for computing power is reflected in its staggering budget for computing operational expenditures, which is projected to exceed USD 450 billion from 2024 to 2030. The company has also inked computing power leasing agreements with multiple entities, and the development of next-generation technologies similarly necessitates massive computing power support. OpenAI's meteoric rise has compelled competitors to ramp up their investments, thereby further solidifying the sustainability of the capital expenditure cycle. Recently, global semiconductor manufacturers have been intensifying their efforts to expand AI chip capacity deployment (here, "deployment" is used to convey the strategic planning and implementation of capacity expansion), and tech giants like Google and Meta have also embarked on relevant investment initiatives, further corroborating Barclays Bank's assessment.