Ads_970x250

Anthropic Weighs $6 Billion Decart Deal

The Claude maker is in early talks to buy the Nvidia-backed startup as it seeks more efficient ways to expand AI computing capacity.

Topics

  • Anthropic is in talks to acquire Nvidia-backed AI startup Decart in a deal that would be worth about $6 billion, Bloomberg reported, as the Claude maker races to expand the computing capacity behind its AI models. 

    The negotiations are at an early stage and could still fall apart. If completed, Decart’s employees would join Anthropic’s inference and performance organization.

    The interest in Decart comes as Anthropic looks for ways to increase the efficiency and capacity of the computing infrastructure behind Claude while preparing for a public listing. The company confidentially filed for a US initial public offering in June. 

    Decart develops AI infrastructure and optimization technology designed to improve AI performance across different processors. Its technology could help Anthropic handle rising demand without relying only on adding more computing hardware.

    The startup also builds AI models. Its Lucy technology can transform live video in real time, while Oasis generates interactive simulated environments that can be used to train and test robotics and autonomous vehicles.

    Decart raised $300 million in May in a funding round led by Radical Ventures, with Nvidia joining as a new investor, according to Reuters.

    The talks also follow Anthropic’s push deeper into the hardware stack. The company recently began hiring engineers with chip and software expertise to work on jointly designing AI models and custom hardware.

    Last week, it said it was building an in-house chip-design team and hiring engineers to jointly design custom chips and AI models aimed at making Claude faster and more efficient. The company said it would continue using hardware from Amazon Web Services, Google, Nvidia and AMD as part of a multi-chip strategy.

    Buying Decart would give Anthropic both infrastructure expertise and a team already working on optimizing AI workloads across different types of hardware, as competition among AI developers increasingly extends from models into the computing systems that run them.

    Topics

    More Like This

    You must to post a comment.

    First time here? : Comment on articles and get access to many more articles.