{"id":3907,"date":"2024-02-14T20:53:12","date_gmt":"2024-02-14T12:53:12","guid":{"rendered":"https:\/\/www.1ai.net\/?p=3907"},"modified":"2024-02-14T20:53:12","modified_gmt":"2024-02-14T12:53:12","slug":"ai-%e8%8a%af%e7%89%87%e9%9c%80%e6%b1%82%e6%bf%80%e5%a2%9e%ef%bc%8chbm-%e5%86%85%e5%ad%98%e4%bb%b7%e6%a0%bc%e6%9a%b4%e6%b6%a8-500","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/3907.html","title":{"rendered":"Demand for AI chips surges, HBM memory prices soar by 500%"},"content":{"rendered":"<p data-vmark=\"7884\">along with<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e4%ba%ba%e5%b7%a5%e6%99%ba%e8%83%bd\" title=\"[View articles tagged with [artificial intelligence]]\" target=\"_blank\" >AI<\/a> The explosive growth of (AI) in 2023 will have a significant impact on <a href=\"https:\/\/www.1ai.net\/en\/tag\/ai%e8%8a%af%e7%89%87\" title=\"[SEES ARTICLES WITH TAGS]\" target=\"_blank\" >AI Chips<\/a>In particular, the demand for HBM memory has surged, causing the average selling price of HBM chips to soar by 500%. This is undoubtedly good news for major manufacturers Micron, Samsung and SK Hynix.<\/p>\n<p data-vmark=\"c6ff\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-3908\" title=\"2587d955-ec02-4490-a727-cd5dd45799f8\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/02\/2587d955-ec02-4490-a727-cd5dd45799f8.jpg\" alt=\"2587d955-ec02-4490-a727-cd5dd45799f8\" width=\"620\" height=\"354\" \/><\/p>\n<p data-vmark=\"2aed\">It is understood that HBM chips are an important part of AI GPUs. For example, AMD and NVIDIA&#039;s flagship AI GPUs use the most advanced HBM memory. Market research firm Yole Group predicts that from 2023 to 2028, HBM supply will grow at a compound annual growth rate of 45%, and because production capacity is difficult to increase quickly to meet the crazy demand, HBM prices will remain high &quot;for some time.&quot;<\/p>\n<p data-vmark=\"d7fa\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-3909\" title=\"f79d827a-d2fb-465a-8b72-c9ffeefbc4e3\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/02\/f79d827a-d2fb-465a-8b72-c9ffeefbc4e3.png\" alt=\"f79d827a-d2fb-465a-8b72-c9ffeefbc4e3\" width=\"1280\" height=\"720\" \/><\/p>\n<p data-vmark=\"63b2\">Samsung and SK Hynix are the dominant players in the HBM market, with the two Korean companies dominating the 90% market share, leaving Micron with only scraps. It is worth noting that SK Hynix and TSMC are joining forces, which could bring some interesting changes in the coming months and years.<\/p>\n<p data-vmark=\"3e4c\">SK Hynix recently revealed that its next-generation HBM4 memory will be mass-produced in 2026 and applied to the next-generation AI GPU after NVIDIA&#039;s upcoming Blackwell B100. As a key component of AI GPU, HBM not only needs to be fast, but also needs larger capacity.<\/p>","protected":false},"excerpt":{"rendered":"<p>With the explosive growth of AI in 2023, demand for the memory of AI chips, particularly HBMs, surged, resulting in an increase in the average selling price of HBM chips by 500%. It's certainly good news for the main producers, Mizuki, Samsung and SK Hercules. HBM chips are known to be an important component of AI GPU, such as the flagship AMD and NVIDIA AI GPU, which use state-of-the-art HBM memory. Market research firm Yole Group predicts that from 2023 to 2028, HBM supply will grow at a compound annual growth rate of 45%, while HBM prices will be difficult to upgrade rapidly to meet the insane demand<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[1099,204],"collection":[],"class_list":["post-3907","post","type-post","status-publish","format-standard","hentry","category-news","tag-ai","tag-204"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/3907","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=3907"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/3907\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=3907"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=3907"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=3907"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=3907"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}