[AI Computing Power Delivery Bottlenecks Emerging? Microsoft Cuts $5 Billion in Capital Expenditure Due to Data Center Delays; SMCI Raises Revenue Forecast] BlockBeats reported on November 7th, citing The Information, that Microsoft reduced its third-quarter capital expenditure by approximately $5 billion from previous plans due to delays in the delivery of third-party data centers it planned to lease. Sources familiar with the matter stated that delays in the Iren data center project in Childress, Texas, led Microsoft to postpone some spending. Microsoft had previously announced plans to invest $9.7 billion over five years to lease servers equipped with Nvidia GPUs from Iren. Power supply and grid access are also bottlenecks in computing power delivery. Last month, Microsoft withdrew from a planned UK data center project leased from Nscale due to grid access delays. Because its self-built data centers cannot meet demand in a timely manner, Microsoft is simultaneously leasing computing power from multiple cloud service providers. Meanwhile, AI server supplier Supermicro Computer Inc. (SMCI) continued to raise its earnings forecast. The company's first-quarter revenue reached $19.95 billion, exceeding analysts' expectations of $16.39 billion, and it projects second-quarter revenue of $19.5 billion to $21.5 billion. The revenue forecast for fiscal year 2027 has also been revised upwards from $65 billion to $72 billion to $72 billion to $82 billion. However, the growth in server sales does not necessarily mean that computing power has been actually put into use. Bernstein points out that assessing AI computing power demand cannot be based solely on contract amounts or megawatt scale; it also requires consideration of contract cancelability, duration, and actual deployment time. Some data center projects may be delivered in phases, spanning months or even years; furthermore, chip procurement agreements and data center leasing contracts may correspond to the same batch of computing power, and if included repeatedly, the actual new computing power demand may be overestimated. This means that the bottleneck in the AI industry chain is extending from chip supply to data center construction, power supply, and computing power deployment. The time lag between server order growth and final computing power delivery warrants attention.