OpenAI reportedly planning to build its first AI chip in 2026

[ { "selector": "#anim-7cd90e4c-071a-4a0c-b8d6-675cb943bdb5", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-3a5ab03b-20f5-4cbe-a9ce-8233dbc85c97", "keyframes": { "transform": [ "translate3d(99.73404%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-560b58f5-c8de-449c-89df-b53b6898d5bb", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-65889cd4-c01a-4283-a372-3992f3f7e572", "keyframes": { "transform": [ "translate3d(-115.63519%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-26f70b21-104e-41ab-a5fb-31428cb838d1", "keyframes": { "opacity": [0, 1] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-9f75e1e1-9cc8-40b6-bbfd-8847fd8274a3", "keyframes": { "transform": [ "translate3d(-115.68628%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] OpenAI reportedly planning to build its first AI chip in 2026 | TechCrunch

OpenAI's Chip Plans

[ { "selector": "#anim-c8d1a841-a311-4c4f-806e-d69822a02e92", "keyframes": { "opacity": [0, 1] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-2605adb9-01dd-44c2-aef6-ffa3f4b9ace7", "keyframes": { "transform": [ "translate3d(-115.68628%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-88c58604-2e20-4602-9d3a-ba0eb89fc302", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-ca8900c7-0b57-43c5-951e-d8685e82e1cc", "keyframes": { "transform": [ "translate3d(99.73404%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-3714bf94-1782-43cb-909d-34259ac82a9e", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-09757cd1-2f17-424f-b571-d42daf1f1b34", "keyframes": { "transform": [ "translate3d(-115.63519%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] OpenAI reportedly working with TSMC and Broadcom to build an in-house AI chip, expected in 2026.

Chip Diversification

[ { "selector": "#anim-31d5e363-1128-4f97-9189-cdb3ea3fea13", "keyframes": { "opacity": [0, 1] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-b4c7eb7b-8bda-4f41-bd68-b848e9becee6", "keyframes": { "transform": [ "translate3d(-115.68628%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-cdb84cda-77c9-4875-a0e4-8edc7dbf5cab", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-e66cba01-2977-4a6f-826e-a56e7c9afa7a", "keyframes": { "transform": [ "translate3d(99.73404%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-0bb73c8b-278b-4d52-b439-cde61c8554d6", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-ac7ffb4f-bdda-4229-91a3-65ee4b10fad0", "keyframes": { "transform": [ "translate3d(-115.63519%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] OpenAI plans to use AMD chips alongside Nvidia GPUs for AI model training. Download App

Abandoned Foundry Ambitions

[ { "selector": "#anim-cb2a3ea6-9668-4f83-8fa2-d5743e83bd5f", "keyframes": { "opacity": [0, 1] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-fe1d6ee2-a2ca-4d05-a285-5764fcb8bb22", "keyframes": { "transform": [ "translate3d(-115.14196%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-17789378-71d3-49d5-8ceb-3e4e020bc07e", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-137bd825-5513-4b20-95d2-3198612d632e", "keyframes": { "transform": [ "translate3d(99.73404%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-26cbe6da-cae2-4609-8627-1e79e083a1d3", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-fb9b2ad1-a107-40db-97d3-67fc3eee93de", "keyframes": { "transform": [ "translate3d(-115.63519%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] OpenAI has abandoned plans to establish its own chip manufacturing network. Download App

Focus on In-House Design

[ { "selector": "#anim-d419589e-11af-4b91-a94a-41b9e82a8d61", "keyframes": { "opacity": [0, 1] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-7759b3a0-a447-4729-bec9-3dfdcc4e1a71", "keyframes": { "transform": [ "translate3d(-114.8265%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 300, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-38a817e6-7e22-47fc-b969-198470908eb8", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-96004321-8a80-4ad5-ade4-9d1845ff0976", "keyframes": { "transform": [ "translate3d(99.73404%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-baf0cd90-3d7a-457c-9a82-4d14017d0aa8", "keyframes": { "opacity": [0, 1] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] [ { "selector": "#anim-f0954e7c-523d-4a4b-a4cc-ce0b9c5be519", "keyframes": { "transform": [ "translate3d(-115.63519%, 0px, 0)", "translate3d(0px, 0px, 0)" ] }, "delay": 0, "duration": 600, "easing": "cubic-bezier(0.2, 0.6, 0.0, 1)", "fill": "both" } ] OpenAI will now focus on in-house chip design rather than manufacturing. Download App