1、Welcome to an Omdia Analyst Summit TimeSessionSpeakers11:00 11:45Where is AI really headed?AI的发展方向在哪里？VladGalabov11:45 12:30 Data center rack architecture and power density out to 20302030年前的数据中心机柜架构与功率密度Manoj Sukumaran12:3014:00Break茶歇14:00 15:00Powering the AI datacenternow and in 2030为当下和2030年的数据

2、中心供电Shen Wang15:0015:30Break茶歇15:30 16:30Cooling the AI data centernow and in 2030为当下和2030年的数据中心散热Jessica Nian16:30 17:00Wrap up and audience Q&A总结与问答环节Vlad GalabovOmdia Analyst Summit:Where is AI really headed?Copyright 2025 TechTarget,Inc.or its subsidiaries.All rights reserved.Where is AI really

3、headed?Vlad GalabovSenior Director,Enterprise Infrastructure ResearchCopyright 2025 TechTarget,Inc.or its subsidiaries.All rights reserved.FundamentalsCopyright 2025 TechTarget,Inc.or its subsidiaries.All rights reserved.Page 5More compute is fundamental for the formation of the AI economyAI usefuln

4、ess and performance will only improve at the expense of power Usefulness Improving productivity gains will accelerate end user adoption Demand Vendor companies will/should not risk getting frozen out of the AI economy Supply Data center investment will not be constrained by cash Cash access Data cen

5、ter investment will not be constrained by power Power access Copyright 2025 TechTarget,Inc.or its subsidiaries.All rights reserved.Page 6More compute is fundamental for the formation of the AI economyAI usefulness and performance will only improve at the expense of power Usefulness Improving product

6、ivity gains will accelerate end user adoption Demand Vendor companies will/should not risk getting frozen out of the AI economy Supply Data center investment will not be constrained by cash Cash access Data center investment will not be constrained by power Power access AI training compute requireme