Anthropic may be preparing to take a much bigger role in the technology powering its own AI models. The company has reportedly held talks to acquire chip startup MatX in a deal that could have been worth around $7 billion, highlighting how urgently AI companies are trying to secure their own hardware advantage.
• Anthropic reportedly discussed a deal valued at about $7 billion
• MatX was founded by former Google TPU engineers
• The talks have reportedly shifted toward a possible partnership
The discussions have since evolved into a potential partnership, according to sources familiar with the matter. MatX is now reportedly seeking fresh funding at a valuation of about $4 billion, while Anthropic continues building an internal team focused on custom silicon. The move could give the AI company greater control over the hardware needed to train and operate increasingly demanding models.
• Anthropic is expanding its in-house chip team
• MatX is reportedly seeking new investment
• Custom hardware could reduce long-term computing costs
Designing advanced AI chips is a costly and lengthy undertaking, making experienced engineers particularly valuable. MatX has been developing technology aimed at training large AI models, while Anthropic has been recruiting veterans from major chip and AI companies to accelerate its own hardware ambitions. The company is expected to maintain a multi-vendor strategy while developing its own processors.
• MatX has expertise in AI model training hardware
• Anthropic has recruited experienced chip engineers
• Nvidia, Google and other suppliers remain part of its strategy
Anthropic’s push comes as demand for AI computing continues to soar. The company is committing tens of billions of dollars toward computing capacity, including major arrangements involving Google’s AI chips and cloud infrastructure, while also securing enormous amounts of capacity from other providers. Building custom processors could eventually give Anthropic more control over performance, energy consumption and costs.
• AI computing demand is driving massive infrastructure spending
• Custom chips could improve efficiency and reduce dependence on suppliers
• Anthropic is still investing heavily in external computing resources
The bigger story may be the shift happening across the AI industry. Anthropic, OpenAI, Google and Amazon are increasingly developing chips tailored to their own workloads as access to high-end processors becomes a strategic advantage. If Anthropic succeeds, its hardware could become just as important to Claude’s future as the AI models themselves.
• Major AI companies are racing to develop custom chips
• Hardware shortages are increasing the pressure to diversify
• Anthropic could eventually control more of the technology behind Claude
Via: Reuters
