OpenAI Inks $38 Billion Deal With Amazon

<div class&equals;"&commat;md&colon;col-span-6 &commat;md&colon;col-start-4 col-span-12 max-w-none &lbrack;&amp&semi;&colon;not&lpar;&colon;first-child&rpar;&rsqb;&colon;mt-sm">&NewLine;<p class&equals;"mb-sm last&colon;mb-0">Amazon Web Services &lpar;AWS&rpar; and OpenAI have formed a multi-year&comma; strategic partnership that provides AWS’s infrastructure to run and scale OpenAI’s core artificial intelligence &lpar;AI&rpar; workloads starting immediately&period; Under this new &dollar;38 billion agreement&comma; which will have continued growth over the next seven years&comma; OpenAI is accessing AWS compute comprising hundreds of thousands of state-of-the-art NVIDIA GPUs&comma; with the ability to expand to tens of millions of CPUs to rapidly scale agentic workloads&period; AWS has unusual experience running large-scale AI infrastructure securely&comma; reliably&comma; and at scale–with clusters topping 500K chips&period; AWS&&num;8217&semi;s leadership in cloud infrastructure combined with OpenAI&&num;8217&semi;s pioneering advancements in generative AI will help millions of users continue to get value from ChatGPT&period;<&sol;p>&NewLine;<&sol;div>&NewLine;<div class&equals;"&commat;md&colon;col-span-6 &commat;md&colon;col-start-4 col-span-12 max-w-none &lbrack;&amp&semi;&colon;not&lpar;&colon;first-child&rpar;&rsqb;&colon;mt-sm">&NewLine;<p class&equals;"mb-sm last&colon;mb-0">The rapid advancement of AI technology has created unprecedented demand for computing power&period; As frontier model providers seek to push their models to new heights of intelligence&comma; they are increasingly turning to AWS due to the performance&comma; scale&comma; and security they can achieve&period; OpenAI will immediately start utilizing AWS compute as part of this partnership&comma; with all capacity targeted to be deployed before the end of 2026&comma; and the ability to expand further into 2027 and beyond&period;<&sol;p>&NewLine;<&sol;div>&NewLine;<div class&equals;"&commat;md&colon;col-span-6 &commat;md&colon;col-start-4 col-span-12 max-w-none &lbrack;&amp&semi;&colon;not&lpar;&colon;first-child&rpar;&rsqb;&colon;mt-sm">&NewLine;<p class&equals;"mb-sm last&colon;mb-0">The infrastructure deployment that AWS is building for OpenAI features a sophisticated architectural design optimized for maximum AI processing efficiency and performance&period; Clustering the NVIDIA GPUs—both GB200s and GB300s—via Amazon EC2 UltraServers on the same network enables low-latency performance across interconnected systems&comma; allowing OpenAI to efficiently run workloads with optimal performance&period; The clusters are designed to support various workloads&comma; from serving inference for ChatGPT to training next generation models&comma; with the flexibility to adapt to OpenAI&&num;8217&semi;s evolving needs&period;<&sol;p>&NewLine;<&sol;div>&NewLine;<div class&equals;"&commat;md&colon;col-span-6 &commat;md&colon;col-start-4 col-span-12 max-w-none &lbrack;&amp&semi;&colon;not&lpar;&colon;first-child&rpar;&rsqb;&colon;mt-sm">&NewLine;<p class&equals;"mb-sm last&colon;mb-0">&OpenCurlyDoubleQuote;Scaling frontier AI requires massive&comma; reliable compute&comma;&&num;8221&semi; said OpenAI co-founder and CEO Sam Altman&period; &OpenCurlyDoubleQuote;Our partnership with AWS strengthens the broad compute ecosystem that will power this next era and bring advanced AI to everyone&period;”<&sol;p>&NewLine;<&sol;div>&NewLine;<div class&equals;"&commat;md&colon;col-span-6 &commat;md&colon;col-start-4 col-span-12 max-w-none &lbrack;&amp&semi;&colon;not&lpar;&colon;first-child&rpar;&rsqb;&colon;mt-sm">&NewLine;<p class&equals;"mb-sm last&colon;mb-0">&OpenCurlyDoubleQuote;As OpenAI continues to push the boundaries of what&&num;8217&semi;s possible&comma; AWS&&num;8217&semi;s best-in-class infrastructure will serve as a backbone for their AI ambitions&comma;” said Matt Garman&comma; CEO of AWS&period; &OpenCurlyDoubleQuote;The breadth and immediate availability of optimized compute demonstrates why AWS is uniquely positioned to support OpenAI&&num;8217&semi;s vast AI workloads&period;&&num;8221&semi;<&sol;p>&NewLine;<&sol;div>&NewLine;<div class&equals;"&commat;md&colon;col-span-6 &commat;md&colon;col-start-4 col-span-12 max-w-none &lbrack;&amp&semi;&colon;not&lpar;&colon;first-child&rpar;&rsqb;&colon;mt-sm">&NewLine;<p class&equals;"mb-sm last&colon;mb-0">This news continues the companies’ work together to provide cutting-edge AI technology to benefit organizations worldwide&period; Earlier this year&comma; OpenAI open weight foundation models became available on Amazon Bedrock&comma; bringing these additional model options to millions of customers on AWS&period; OpenAI has quickly become one of the most popular publicly available model providers in Amazon Bedrock with thousands of customers—including Bystreet&comma; Comscore&comma; Peloton&comma; Thomson Reuters&comma; Triomics&comma; and Verana Health—working with their models for agentic workflows&comma; coding&comma; scientific analysis&comma; mathematical problem-solving&comma; and more&period;<&sol;p>&NewLine;<&sol;div>&NewLine;