Meta signs deal for millions of Amazon AI CPUs

Amazon has secured a significant partnership with Meta, once again relying on its own custom-designed chips. Meta has agreed to deploy millions of AWS Graviton chips to meet its expanding AI demands, Amazon confirmed on Friday.
Note that AWS Graviton is an ARM-based CPU (central processing unit, designed for general computing), not a GPU (graphics processing unit).
While GPUs remain the preferred chip for training large models, once those models are trained, AI agents built on top of them are driving a shift in the type of chip required. These agents generate compute-heavy workloads such as real-time reasoning, code writing, search, and coordinating multi-step tasks. AWS says its latest Graviton version was purpose-built to handle AI-related computing needs.
This agreement redirects more of Meta's spending to AWS rather than to competitors like Google Cloud. In August last year, Meta signed a six-year, $10 billion deal with Google Cloud, though Meta had previously been predominantly an AWS customer that also utilized Microsoft Azure.
We noticed that AWS timed the announcement of this deal to coincide with the conclusion of the Google Cloud Next conference, almost like a veiled jab at its cloud rival. Google, of course, also develops its own custom AI chips and unveiled new versions during the event.
Admittedly, Amazon also manufactures its own AI GPU: the Trainium, which, despite its name, is used for both training and inference — the phase after a model is trained, when it actively processes prompts.
However, Anthropic had already secured a deal announced earlier this month that reserved many of those chips for years ahead. The company behind Claude agreed to spend $100 billion over ten years to run its workloads on AWS, with a particular emphasis on Trainium, while Amazon agreed to invest an additional $5 billion (bringing total investment to $13 billion) into Anthropic in return.
Ultimately, the Meta deal allows Amazon to highlight a major AI customer as a proof point for its own CPUs. These chips compete with Nvidia's new Vera CPU, which is also ARM-based and designed for AI agent workloads. The key difference is that Nvidia sells its chips and AI systems to enterprises and cloud providers, including AWS, while AWS only offers access to its chips through its cloud service.
Earlier this month, Amazon CEO Andy Jassy criticized Nvidia and Intel in his annual shareholder letter, stating that enterprises demand better price-performance ratios for AI and that he plans to win deals on that basis. This also means the pressure on Amazon's internal chip development team to deliver has never been greater. We visited that team last month during an exclusive tour of their lab.
Related article
Amazon Launches Free Alexa+ on Fire TV Without Prime Membership
AI is making its way into your television, whether you’re ready for it or not.On Wednesday, Amazon announced that its AI assistant, Alexa+, will be available on all compatible Fire TV devices in the U.S. at no cost, regardless of Prime subscription s
Amazon introduces Alexa for Shopping while pushing Rufus to the background
Amazon has launched Alexa for Shopping, merging its Rufus shopping chatbot with Alexa+ across the app, website, and Echo Show devices.The assistant answers product queries, compares items, tracks prices, and supports shopping reminders. It also handl
Frontier AI Labs Refuse to Disclose Containment Strategies for Rogue Models
Recent research indicates that very few leading AI laboratories have published or demonstrated containment response plans. A containment plan defines the procedures for when an AI system attempts to subvert human control, specifying which access righ
Related Special Topic Recommendations
Comments (0)
0/500

Amazon has secured a significant partnership with Meta, once again relying on its own custom-designed chips. Meta has agreed to deploy millions of AWS Graviton chips to meet its expanding AI demands, Amazon confirmed on Friday.
Note that AWS Graviton is an ARM-based CPU (central processing unit, designed for general computing), not a GPU (graphics processing unit).
While GPUs remain the preferred chip for training large models, once those models are trained, AI agents built on top of them are driving a shift in the type of chip required. These agents generate compute-heavy workloads such as real-time reasoning, code writing, search, and coordinating multi-step tasks. AWS says its latest Graviton version was purpose-built to handle AI-related computing needs.
This agreement redirects more of Meta's spending to AWS rather than to competitors like Google Cloud. In August last year, Meta signed a six-year, $10 billion deal with Google Cloud, though Meta had previously been predominantly an AWS customer that also utilized Microsoft Azure.
We noticed that AWS timed the announcement of this deal to coincide with the conclusion of the Google Cloud Next conference, almost like a veiled jab at its cloud rival. Google, of course, also develops its own custom AI chips and unveiled new versions during the event.
Admittedly, Amazon also manufactures its own AI GPU: the Trainium, which, despite its name, is used for both training and inference — the phase after a model is trained, when it actively processes prompts.
However, Anthropic had already secured a deal announced earlier this month that reserved many of those chips for years ahead. The company behind Claude agreed to spend $100 billion over ten years to run its workloads on AWS, with a particular emphasis on Trainium, while Amazon agreed to invest an additional $5 billion (bringing total investment to $13 billion) into Anthropic in return.
Ultimately, the Meta deal allows Amazon to highlight a major AI customer as a proof point for its own CPUs. These chips compete with Nvidia's new Vera CPU, which is also ARM-based and designed for AI agent workloads. The key difference is that Nvidia sells its chips and AI systems to enterprises and cloud providers, including AWS, while AWS only offers access to its chips through its cloud service.
Earlier this month, Amazon CEO Andy Jassy criticized Nvidia and Intel in his annual shareholder letter, stating that enterprises demand better price-performance ratios for AI and that he plans to win deals on that basis. This also means the pressure on Amazon's internal chip development team to deliver has never been greater. We visited that team last month during an exclusive tour of their lab.
Amazon Launches Free Alexa+ on Fire TV Without Prime Membership
AI is making its way into your television, whether you’re ready for it or not.On Wednesday, Amazon announced that its AI assistant, Alexa+, will be available on all compatible Fire TV devices in the U.S. at no cost, regardless of Prime subscription s
Frontier AI Labs Refuse to Disclose Containment Strategies for Rogue Models
Recent research indicates that very few leading AI laboratories have published or demonstrated containment response plans. A containment plan defines the procedures for when an AI system attempts to subvert human control, specifying which access righ





Home






