Infrastructure

On September 14, Inweida CEO Hoang In-hoon reaffirmed the global AI infrastructure expenditure forecast to reach $3 trillion to $4 trillion by 2030 in Goldman Sachs on September 10. In his view, the calculations were moving from retrieving existing information to producing answers in real time, slowing down the Jamour law and requiring integrated and synergistic design of chips, whole machines and data centres. Wong In-hoon said that the infrastructure supply constraints are both upstream and downstream..

ON 5 JUNE, THE UNITED NATIONS UNIVERSITY RELEASED A REPORT YESTERDAY BY THE INSTITUTE FOR WATER, ENVIRONMENT AND HEALTH THAT THE RAPID EXPANSION OF ARTIFICIAL INTELLIGENCE INFRASTRUCTURE IS SIGNIFICANTLY INCREASING ENERGY, WATER AND LAND NEEDS. THE REPORT ESTIMATES THAT THE GLOBAL DATA CENTRE WILL CONSUME 448 WATT-HOURS IN 2025; BY 2030, THE DATA CENTRE ' S ELECTRICITY DEMAND MAY INCREASE TO 945 WATT-HOURS (1 WATT-HOUR = 1 BILLION DEGREES OF ELECTRICITY), REPRESENTING APPROXIMATELY 31 TP3T OF THE TOTAL GLOBAL ELECTRICITY CONSUMPTION. THE REPORT LISTS “DEBATE” AS THE MAIN SOURCE OF ENERGY CONSUMPTION. RESEARCH SHOWS THAT THE REASONING CHAIN ACCOUNTS FOR 801 TP3T TO 90..

ON MARCH 11, THE NEWS CAME OUT YESTERDAY, WHEN YOUNG WEIDA PUBLISHED A SIGNED ARTICLE BY CEO WONG IN-HOON PROPOSING AN AI "FIVE LAYERS OF CAKE" ARCHITECTURE MODEL AND CLEARLY CHARACTERIZING AI AS A MODERN INFRASTRUCTURE OF EQUAL IMPORTANCE TO ELECTRICITY. IN HIS ARTICLE, WONG IN-HOON NOTED THAT COMPUTING TECHNOLOGY WAS SHIFTING FROM TRADITIONAL PREFABRICATED SOFTWARE TO REAL-TIME-GENERATED INTELLIGENCE AND THAT THE EVOLUTION OF THIS BOTTOM LOGIC REQUIRED A COMPLETE RE-ENGINEERING OF THE ENTIRE COMPUTING ARCHITECTURE. HE DECOMPOSED THE AI ARCHITECTURE FROM THE BOTTOM UP TO FIVE LEVELS: ENERGY, CHIPS, INFRASTRUCTURE, MODELS AND APPLICATIONS. IT EMPHASIZES THAT ENERGY IS THE PRIMARY CONSTRAINT ON THE SCALE OF OUTPUT OF SMART SYSTEMS..