{"id":23906,"date":"2026-09-09T14:30:00","date_gmt":"2026-09-09T14:30:00","guid":{"rendered":"https:\/\/scannn.com\/openai-says-its-next-generation-processors-could-be-made-at-samsung-double-sourcing-with-tsmc-hints-at-massive-volume-requirements-updated\/"},"modified":"2026-09-09T14:30:00","modified_gmt":"2026-09-09T14:30:00","slug":"openai-says-its-next-generation-processors-could-be-made-at-samsung-double-sourcing-with-tsmc-hints-at-massive-volume-requirements-updated","status":"publish","type":"post","link":"https:\/\/scannn.com\/lv\/openai-says-its-next-generation-processors-could-be-made-at-samsung-double-sourcing-with-tsmc-hints-at-massive-volume-requirements-updated\/","title":{"rendered":"OpenAI says its next-generation processors could be made at Samsung \u2014 double-sourcing with TSMC hints at massive volume requirements [Updated]"},"content":{"rendered":"\n<div id=\"article-body\">\n<div class=\"mb-6\" data-component-name=\"Elk:InfoBlock\" id=\"elk-b9e901fa-ade2-11f1-b1ad-2debf1ca5c0f\">\n<div class=\"[&amp;_a]:text-[#2f6e91] [&amp;_a]:visited:text-[#09c] p-5 bg-[#F5F5F5] text-sm text-[#333]\">\n<p class=\"infoUpdate-log tracking-[.02em] pt-2 text-base font-normal\"><strong>Update &#8211; Friday 11 Sept, 9am ET: <\/strong>An OpenAI spokesperson has given <em>Tom&#8217;s Hardware<\/em> the following statement: &#8220;The reference to expanding our collaboration with Samsung was made in the context of our existing relationship, including Samsung Electronics\u2019 adoption of ChatGPT and the LOI announced last October. We have nothing new to announce. We did not announce plans to manufacture chips through Samsung.&#8221;<br \/>The original story remains below as published. <\/p>\n<\/div>\n<\/div>\n<p id=\"elk-d7e5c768-ac54-11f1-8178-79c68a3a1816\">OpenAI is expanding its relationship with Samsung beyond memory supply and enterprise software as the AI giant plans to outsource production of at least some of its processors to Samsung Foundry, Harrison Kim, General Manager of OpenAI Korea, revealed this week. If the information is accurate, then OpenAI will source its AI accelerators from both TSMC and Samsung Foundry, which suggests massive volume requirements.<\/p>\n<div class=\"fancy-box\" id=\"elk-6cbeea4a-ac55-11f1-9603-bb94aebe1da0\">\n<p>Go deeper with TH Premium: CPU<\/p>\n<div class=\"fancy_box_body\">\n<figure class=\"van-image-figure \">\n<div class=\"image-full-width-wrapper\">\n<div class=\"image-widthsetter\">\n<p class=\"vanilla-image-block\" style=\"padding-top:56.25%;\"> <picture data-new-v2-image=\"true\"><source type=\"image\/webp\" srcset=\"https:\/\/cdn.mos.cms.futurecdn.net\/Xh2MupWrRjJPiLLuopmKRB-1200-80.jpg.webp 1200w, https:\/\/cdn.mos.cms.futurecdn.net\/Xh2MupWrRjJPiLLuopmKRB-1024-80.jpg.webp 1024w, https:\/\/cdn.mos.cms.futurecdn.net\/Xh2MupWrRjJPiLLuopmKRB-970-80.jpg.webp 970w, https:\/\/cdn.mos.cms.futurecdn.net\/Xh2MupWrRjJPiLLuopmKRB-650-80.jpg.webp 650w, https:\/\/cdn.mos.cms.futurecdn.net\/Xh2MupWrRjJPiLLuopmKRB-480-80.jpg.webp 480w, https:\/\/cdn.mos.cms.futurecdn.net\/Xh2MupWrRjJPiLLuopmKRB-320-80.jpg.webp 320w\" sizes=\"(min-width: 1000px) 970px, calc(100vw - 40px)\"\/>\n<\/picture><\/p>\n<\/div>\n<\/div><figcaption itemprop=\"caption description\" class=\"\"><span class=\"credit\" itemprop=\"copyrightHolder\">(Image credit: Tom&#8217;s Hardware)<\/span><\/figcaption><\/figure>\n<\/div>\n<\/div>\n<p id=\"elk-6cbeeacc-ac55-11f1-abda-ef771cfef533\">&#8220;One of the areas where we have made the most progress and gained the most recognition with Samsung Electronics is our joint production and \u200bresearch on the next-generation chips we are developing,&#8221; said Harrison Kim, General Manager of OpenAI Korea, at \u200ba press conference in Seoul, <a data-analytics-id=\"inline-link\" href=\"https:\/\/www.reuters.com\/world\/asia-pacific\/openai-says-working-with-samsung-next-generation-chips-deepening-cooperation-2026-09-09\/\" data-url=\"https:\/\/www.reuters.com\/world\/asia-pacific\/openai-says-working-with-samsung-next-generation-chips-deepening-cooperation-2026-09-09\/\" target=\"_blank\" referrerpolicy=\"no-referrer-when-downgrade\" data-hl-processed=\"none\" data-article-category=\"tech-industry\" data-mrf-recirculation=\"inline-link\"><em>Reuters<\/em><\/a> reports.<\/p>\n<p><a id=\"elk-seasonal\"\/><\/p>\n<aside data-article-category=\"tech-industry\" data-block-type=\"embed\" data-render-type=\"fte\" data-skip=\"dealsy\" data-widget-type=\"seasonal\" class=\"ecom-root\"\/>\n<p id=\"elk-6cbeeacc-ac55-11f1-abda-ef771cfef533-1\">OpenAI already has its own AI ASIC program that relies on Broadcom&#8217;s design services as well as TSMC&#8217;s wafer processing and advanced packaging services. So far, the company has introduced its <a data-analytics-id=\"inline-link\" href=\"https:\/\/www.tomshardware.com\/tech-industry\/artificial-intelligence\/hot-chips-2026-openais-jalapeno-ai-asic-unpacked-accelerator-developed-using-ai-achieves-efficiency-and-throughput-gains-against-power-hungry-blackwell\" data-url=\"https:\/\/www.tomshardware.com\/tech-industry\/artificial-intelligence\/hot-chips-2026-openais-jalapeno-ai-asic-unpacked-accelerator-developed-using-ai-achieves-efficiency-and-throughput-gains-against-power-hungry-blackwell\" data-hl-processed=\"none\" data-article-category=\"tech-industry\" data-mrf-recirculation=\"inline-link\" data-before-rewrite-localise=\"https:\/\/www.tomshardware.com\/tech-industry\/artificial-intelligence\/hot-chips-2026-openais-jalapeno-ai-asic-unpacked-accelerator-developed-using-ai-achieves-efficiency-and-throughput-gains-against-power-hungry-blackwell\">first inference AI accelerator called Jalape\u00f1o<\/a> that was defined by the company&#8217;s engineers, then co-designed with Broadcom, then made by TSMC, all in less than 18 months.<\/p>\n<div class=\"my-6 w-full overflow-hidden rounded-[10px] lg:my-8\" data-component-name=\"JwPlayer:Carousel\" data-jwp-carousel=\"\" data-jwp-carousel-payload=\"{&quot;ids&quot;:{&quot;playerID&quot;:&quot;3ctY47st&quot;,&quot;searchPlaylistID&quot;:&quot;E849Pwt6&quot;,&quot;divID&quot;:&quot;botr_E849Pwt6_3ctY47st_div&quot;,&quot;fallbackPlaylistID&quot;:&quot;E849Pwt6&quot;,&quot;fallbackDivID&quot;:&quot;botr_E849Pwt6_3ctY47st_div&quot;,&quot;key&quot;:&quot;PQRPe9rcDKVpvWTxi78IpFIsjl2YGAdypF5ix2YVG1nfZtwI&quot;,&quot;tintLogo&quot;:false,&quot;useSearchPlaylist&quot;:false,&quot;enabled&quot;:true},&quot;signPostingEnabled&quot;:true,&quot;signPostingLinkEnabled&quot;:true,&quot;waitForAdLoad&quot;:false,&quot;hidePlayerOnDesktop&quot;:false,&quot;hidePlayerOnMobile&quot;:false,&quot;hidePlayerOnTablet&quot;:false}\">\n<div class=\"flex flex-nowrap items-center justify-between gap-3 bg-zinc-900 px-[14px] py-3\" data-jwp-carousel-header=\"\">\n<p><span class=\"inline-flex items-center gap-1.5 text-sm font-article-heading capitalize leading-5 text-white whitespace-nowrap\"><span class=\"jwp-carousel-title-mobile\"\/><span class=\"jwp-carousel-title-desktop\">Latest Videos From<\/span><span class=\"jwp-carousel-brand\">Tom&#8217;s Hardware<\/span><\/span><\/p>\n<\/div>\n<\/div>\n<p>OpenAI did not explain whether Samsung&#8217;s participation concerns a second production source for Jalape\u00f1o, another processor under development by OpenAI, a chip jointly developed by Samsung and OpenAI, or some other aspect of chip production and development. Samsung and SK hynix already supply memory for OpenAI&#8217;s Stargate data center initiative, though joint chip development and production barely have a relation to DRAM supply.<\/p>\n<p>OpenAI&#8217;s 1<sup>st<\/sup> Generation Jalape\u00f1o will unlikely be double-sourced from TSMC and Samsung because the chip is already in mass production at TSMC and OpenAI is talking about &#8216;next-generation chips,&#8217; not the ones that are in mass production at the moment. Furthermore, development of Jalape\u00f1o&#8217;s successor is well underway and is approaching tapeout, which means that its mass production is not far away either. Since OpenAI&#8217;s claim clearly involves &#8216;next-generation chips,&#8217; it is entirely possible that OpenAI will indeed produce its 2<sup>nd<\/sup> Generation inference ASIC at Samsung Foundry.<\/p>\n<aside data-component-name=\"Recirculation:ArticleRiver\" data-recirculation-type=\"inline\" data-mrf-recirculation=\"Trending Bar\" data-nosnippet=\"\" class=\"clear-both pt-2 pb-0 mb-4\">\n        <span class=\"&#10;            flex&#10;            after:content-[''] after:flex-1 after:ml-4 after:my-[0.7rem] after:border-t after:border-solid after:border-t-[#ccc]&#10;            before:content-[''] before:flex-1 before:mr-4 before:my-[0.7rem] before:border-t before:border-solid before:border-t-[#ccc]&#10;            font-article-heading pb-0 text-[length:var(--article-river-title--font-size,1em)] uppercase sm:text-[length:var(--article-river-title--font-size,0.875em)] font-bold&#10;        \"><br \/>\n            You may like<br \/>\n        <\/span><\/p>\n<\/aside>\n<p>Back in late July, Samsung Electronics and Broadcom announced a strategic partnership valued at over $200 billion through 2030 to collaborate on advanced foundry, memory, and packaging technologies for AI infrastructure. Hence, as OpenAI has an agreement with Broadcom to procure 10GW of custom AI accelerators, it will be able to produce these accelerators at both Samsung and TSMC. Of course, if it needs silicon produced at Samsung, and pays Broadcom for appropriate design porting.<\/p>\n<p>Perhaps, OpenAI will take a page from Tesla&#8217;s book and will double-source Jalape\u00f1o&#8217;s successor from TSMC and Samsung to get higher volumes. However, Tesla&#8217;s volume requirements may be different from those of OpenAI.<\/p>\n<div id=\"slice-container-newsletterForm-articleInbodyContent-9bT792FiwXmnt6DfXbhcRg\" class=\"slice-container newsletter-inbodyContent-slice newsletterForm-articleInbodyContent-9bT792FiwXmnt6DfXbhcRg slice-container-newsletterForm\">\n<div data-hydrate=\"true\" class=\"newsletter-form__wrapper newsletter-form__wrapper--inbodyContent\">\n<div class=\"newsletter-form__container\">\n<section class=\"newsletter-form__top-bar\"\/>\n<section class=\"newsletter-form__main-section\">\n<p class=\"newsletter-form__strapline\">Get Tom&#8217;s Hardware&#8217;s best news and in-depth reviews, straight to your inbox.<\/p>\n<\/section>\n<\/div>\n<\/div>\n<\/div>\n<p>Tesla needs extraordinary AI5 volumes because it intends to use the processor across three very different high-volume applications: AI data centers, vehicles, and Optimus robots. Therefore, Tesla could potentially need millions of AI5 chips for cars alone, on top of robots and data-center deployments. Therefore, paying for separate TSMC and Samsung physical implementations gives Tesla not only supply-chain resilience but also aggregate capacity necessary to supply several product categories.<\/p>\n<p>Yet, data center accelerators tend to be vastly more silicon-intensive per unit compared to ASICs for vehicles or robots. If OpenAI\/Broadcom&#8217;s next ASIC is a large leading-edge processor with multiple dies and OpenAI wants gigawatts of these processors, wafer requirements could still become too high for TSMC alone (which is fully booked by the likes of AMD and Nvidia). In that situation, OpenAI may need another foundry to get enough ASICs. Still, we are speculating.<\/p>\n<p><a href=\"https:\/\/news.google.com\/publications\/CAAqLAgKIiZDQklTRmdnTWFoSUtFSFJ2YlhOb1lYSmtkMkZ5WlM1amIyMG9BQVAB\" id=\"elk-d7e5c8b2-ac54-11f1-b667-b15515ebccee\" data-url=\"https:\/\/news.google.com\/publications\/CAAqLAgKIiZDQklTRmdnTWFoSUtFSFJ2YlhOb1lYSmtkMkZ5WlM1amIyMG9BQVAB\" target=\"_blank\" referrerpolicy=\"no-referrer-when-downgrade\" data-hl-processed=\"none\" data-article-category=\"tech-industry\"><\/p>\n<figure class=\"van-image-figure pull-left inline-layout\" data-bordeaux-image-check=\"\">\n<div class=\"image-full-width-wrapper\">\n<div class=\"image-widthsetter\" style=\"max-width:676px;\">\n<p class=\"vanilla-image-block\" style=\"padding-top:31.51%;\"> <picture data-new-v2-image=\"true\"><source type=\"image\/webp\" srcset=\"https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-676-80.png.webp 1200w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-676-80.png.webp 1024w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-676-80.png.webp 970w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-650-80.png.webp 650w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-480-80.png.webp 480w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-320-80.png.webp 320w\" sizes=\"(min-width: 1000px) 970px, calc(100vw - 40px)\"\/><img decoding=\"async\" alt=\"Google Preferred Source\" srcset=\"https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-676-80.png 1200w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-676-80.png 1024w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-676-80.png 970w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-650-80.png 650w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-480-80.png 480w, https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56-320-80.png 320w\" sizes=\"(min-width: 1000px) 970px, calc(100vw - 40px)\" loading=\"lazy\" data-new-v2-image=\"true\" src=\"https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56.png\" data-pin-media=\"https:\/\/cdn.mos.cms.futurecdn.net\/7cUTDmN2PHNRiNBVqbKf56.png\" class=\"rounded-[var(--image--border-radius,0)] pull-left\"\/>\n<\/picture><\/p>\n<\/div>\n<\/div>\n<\/figure>\n<p><\/a><\/p>\n<p id=\"elk-d7e5c934-ac54-11f1-a594-4d5110477fd6\"><em>Follow<\/em><a data-analytics-id=\"inline-link\" href=\"https:\/\/news.google.com\/publications\/CAAqLAgKIiZDQklTRmdnTWFoSUtFSFJ2YlhOb1lYSmtkMkZ5WlM1amIyMG9BQVAB\" target=\"_blank\" data-url=\"https:\/\/news.google.com\/publications\/CAAqLAgKIiZDQklTRmdnTWFoSUtFSFJ2YlhOb1lYSmtkMkZ5WlM1amIyMG9BQVAB\" referrerpolicy=\"no-referrer-when-downgrade\" data-hl-processed=\"none\" data-article-category=\"tech-industry\" data-mrf-recirculation=\"inline-link\"><em> Tom&#8217;s Hardware on Google News<\/em><\/a><em>, or<\/em><a data-analytics-id=\"inline-link\" href=\"https:\/\/google.com\/preferences\/source?q=\" target=\"_blank\" data-url=\"https:\/\/google.com\/preferences\/source?q=\" referrerpolicy=\"no-referrer-when-downgrade\" data-hl-processed=\"none\" data-article-category=\"tech-industry\" data-mrf-recirculation=\"inline-link\"><em> add us as a preferred source<\/em><\/a><em>, to get our latest news, analysis, &amp; reviews in your feeds.<\/em><\/p>\n<\/div>\n<p><a href=\"https:\/\/www.tomshardware.com\/tech-industry\/artificial-intelligence\/openai-says-its-next-generation-processors-could-be-made-at-samsung-double-sourcing-with-tsmc-hints-at-massive-volume-requirements?utm_source=tldrinfosec\">Source link <\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Update &#8211; Friday 11 Sept, 9am ET: An OpenAI spokesperson has given Tom&#8217;s Hardware the following statement: &#8220;The reference to expanding our collaboration with Samsung was made in the context of our existing relationship, including Samsung Electronics\u2019 adoption of ChatGPT and the LOI announced last October. We have nothing new to announce. We did not [&hellip;]<\/p>\n","protected":false},"author":16,"featured_media":23907,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[143],"tags":[],"class_list":["post-23906","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai"],"_links":{"self":[{"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/posts\/23906","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/users\/16"}],"replies":[{"embeddable":true,"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/comments?post=23906"}],"version-history":[{"count":0,"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/posts\/23906\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/media\/23907"}],"wp:attachment":[{"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/media?parent=23906"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/categories?post=23906"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/scannn.com\/lv\/wp-json\/wp\/v2\/tags?post=23906"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}