AMD to Back Anthropic with Up to $5 Billion in AI Infrastructure Deal

AMD is set to invest up to $5 billion in Anthropic, part of a strategic partnership in AI infrastructure. Anthropic will install 2 GW of AMD Instinct MI450 GPUs in AMD Helios rack-scale systems to increase compute capacity for their Claude AI models.

AMD to back Anthropic with up to $5 billion in AI infrastructure deal
AMD to back Anthropic with up to $5 billion in AI infrastructure deal

In a joint announcement made public on July 22nd, Advanced Micro Devices stated its intention to invest as much as $5 billion in Anthropic, an artificial intelligence firm, "in the future". The agreement includes the chipmaker's announcement that Anthropic will use 2 GW of GPUs from the AMD Instinct MI450 Series in AMD Helios rack-scale solutions.

A release states that the first gigawatt, a commonly used metric to describe the capacity of AI data centres, will be deployed in the first half of next year. AMD is excited to expand its collaboration with Anthropic and deploy AMD Helios at gigawatt scale, according to a statement by AMD CEO Lisa Su.

Anthropic Building its Network

Anthropic has announced a series of infrastructure arrangements this year in an effort to enhance its processing capacity, the most recent of which is its agreement with AMD. In April, the business announced that the "inevitable strain" on its infrastructure due to demand for Claude models and goods had an effect on the "reliability and performance" of its services, particularly during busy periods.

Anthropic and Elon Musk's SpaceX reached an agreement in May to utilise SpaceX's Colossus 1 data centre in Memphis, Tennessee, to its full computing capacity. As per SpaceX's prospectus, Anthropic will pay $1.25 billion monthly until May 2029 as part of that arrangement. A multi-gigawatt compute agreement with Broadcom and Google and a multi-billion-dollar deal with Amazon were inked by the firm in April. Meta and Anthropic are in early discussions regarding the leasing of computing capacity.

Claude Code Popularity Chalking the Success of Anthropic

Thanks in large part to the popularity of its artificial intelligence coding assistant, Claude Code, Anthropic has seen meteoric growth this year. May saw the company announce run-rate revenue of nearly $47 billion, up from $10 billion for the whole year prior. In the same month, the firm raised $965 billion in cash and is now preparing for an initial public offering (IPO) that might happen this year.

Anthropic has kept mum about its future plans since June, when it quietly submitted its prospectus to the SEC. The artificial intelligence boom has been very beneficial to AMD as well, as the company produces graphics processing units (GPUs) that are capable of running heavy workloads and training models. Not only does it take on market leader Nvidia head-on, but it has also signed partnerships with other prominent industry figures, such as OpenAI, an ally of Anthropic's.

Under the terms of the agreement, AMD granted OpenAI the right to purchase up to 160 million shares of AMD common stock. Vesting milestones are connected to both the volume of deployments and AMD's share price; these shares are roughly 10% of the firm.