{"id":52028,"date":"2026-04-10T11:38:50","date_gmt":"2026-04-10T03:38:50","guid":{"rendered":"https:\/\/www.1ai.net\/?p=52028"},"modified":"2026-04-10T11:38:50","modified_gmt":"2026-04-10T03:38:50","slug":"google-ceo%ef%bc%9a%e5%86%85%e5%ad%98%e5%b7%b2%e7%bb%8f%e6%88%90%e4%b8%ba%e5%bd%93%e5%89%8d-ai-%e5%9f%ba%e7%a1%80%e8%ae%be%e6%96%bd%e4%b8%ad%e3%80%8c%e6%9c%80%e5%85%b3%e9%94%ae%e7%9a%84%e8%a6%81","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/52028.html","title":{"rendered":"Google CEO: Memory has become one of the \"most critical elements\" of the current AI infrastructure"},"content":{"rendered":"<p>April 10 news.<a href=\"https:\/\/www.1ai.net\/en\/tag\/google\" title=\"[View articles tagged with [Google]]\" target=\"_blank\" >Google<\/a> CEO SANDAL PICHAI<a href=\"https:\/\/www.1ai.net\/en\/tag\/sundar-pichai\" title=\"[See article with [Sundar Pichai] label]\" target=\"_blank\" >Sundar Pichai<\/a>IN A RECENT PUBLIC INTERVIEW, IT WAS STATED THAT THE PHYSICAL BOTTLENECKS IN COMPUTING AND SUPPLY CHAINS WILL BE A KEY FACTOR LIMITING THE PACE OF AI EXPANSION THIS YEAR AND NEXT YEAR\u3002<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-52029\" title=\"bf104decj00td9e3r002bd000ugwm\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2026\/04\/bf104decj00td9e3r002bd000u000gwm.jpg\" alt=\"bf104decj00td9e3r002bd000ugwm\" width=\"1080\" height=\"608\" \/><\/p>\n<p>According to Pichai, although Google\u2019s capital spending this year is expected to be between $175 billion and $185 billion, \u201cmoney does not solve all the problems.\u201d\u3002<\/p>\n<p>He pointed out that even if an enterprise were willing to expand its capital spending to $40 billion, it would not necessarily be able to actually put it into operation, \"because<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e5%86%85%e5%ad%98\" title=\"[See articles containing [RAM] labels]\" target=\"_blank\" >Memory<\/a>There's not enough, there's not enough power.\"\u3002<\/p>\n<p><strong>IN HIS VIEW, MEMORY HAS BECOME ONE OF THE \"MOST CRITICAL ELEMENTS\" OF THE CURRENT AI INFRASTRUCTURE<\/strong>In the short term, it is almost impossible for leading memory producers to achieve a significant increase in production, making supply-side constraints difficult to ease in the coming year or two\u3002<\/p>\n<p>Also, PC Gamer reported that the CEO Michael Dell of Dale Tech Group had predicted that by 2028, total memory needs in the global AI accelerator area would be 625 times as high as in 2022, when attending a series of interviews with the Bank of America's Top CEO Perspectives\u3002<\/p>\n<p>DALE POINTED OUT THAT ONLY HERCULES, SAMSUNG AND MIGWANG CURRENTLY HAVE HBM4 CAPACITY GLOBALLY, WHILE THE CONSTRUCTION OF A NEW DRAM CRYSTAL CIRCLE PLANT TAKES ABOUT FOUR YEARS FROM PLANNING TO PRODUCTION. IN 2023, AT THE LOW END OF THE STORAGE INDUSTRY CYCLE, THE PRODUCTION OF THE THREE MAIN PLANTS WAS SUSPENDED OWING TO SIGNIFICANT LOSSES, WHICH DIRECTLY LED TO A SEVERE SHORTAGE OF SUBSEQUENT CAPACITY RESERVES\u3002<\/p>\n<p>IN HIS VIEW, EVEN IF THE MEMORY PRICE RISES OR THE PROCUREMENT TEMPO IS DELAYED, THE COMMITMENT OF ENTERPRISES AND CLOUD SERVICE PROVIDERS TO THE AI INFRASTRUCTURE WILL NOT WAVER, AND THE RIGID DEMAND FOR THE AI MEMORY BY INDUSTRY WILL FURTHER EXACERBATE SUPPLY AND DEMAND TENSIONS<strong>The shortfall is expected to continue until 2028<\/strong>.<\/p>","protected":false},"excerpt":{"rendered":"<p>On April 10th, Google CEO Sundar Pichai stated in a recent public interview that physical bottlenecks in computing and supply chains will be a key factor limiting the pace of AI expansion this year and next. According to Pichai, although Google\u2019s capital spending this year is projected to be between $175 billion and $185 billion, \u201cmoney does not solve all the problems\u201d. He pointed out that even if enterprises were willing to expand capital spending to $40 billion, they might not be able to actually put it into service, \u201cbecause of insufficient memory and insufficient electricity\u201d. In his view, memory has become one of the \"most critical elements\" of the current AI infrastructure, while leading memory manufacturers are almost impossible in the short term<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[308,1417,8451],"collection":[],"class_list":["post-52028","post","type-post","status-publish","format-standard","hentry","category-news","tag-google","tag-sundar-pichai","tag-8451"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/52028","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=52028"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/52028\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=52028"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=52028"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=52028"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=52028"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}