Anthropic Starts Developing Its Own Chip for AI

Anthropic, one of the leading companies in the AI sector, is forming a dedicated team of engineers to design its own specialized AI chips. According to Business Insider, the creator of the Claude chatbot plans to develop hardware and neural network models together to increase the speed and efficiency of its technologies. The sharp rise in demand for modern technologies has recently forced leaders in the AI market to significantly expand their infrastructure. Techcrunch.com reports .

Last month, The Information reported that Anthropic was considering Samsung as a potential partner for manufacturing its chips. The company’s main motivation for taking this step is reportedly the continuously growing demand for its Claude service. At present, major players in the AI market are competing to secure as much computing power as possible, and Anthropic is no exception.

A Strategic Step Toward Expansion

So far, Anthropic has signed strategic agreements with giants such as AWS, Google, NVIDIA and AMD to meet its computing needs. While these partnerships have given the company access to the necessary hardware, experts say relying solely on external suppliers is not enough to fully meet the current pace of global AI demand. That is why creating a specialized silicon team and developing its own hardware solutions are becoming strategically important.

The company is reportedly hiring qualified engineers with chip design experience for its “specialized silicon team.” According to the requirements listed in the job postings, candidates must have practical skills in designing high-performance chips and integrating them with AI models. Anthropic’s press office has so far declined to officially comment on the information.

A Broader Market Trend

It is worth noting that Anthropic is not the only AI company to decide to develop its own processors. A distinct competitive environment has emerged in the technology market in this area:

• In June this year, OpenAI unveiled a chip called Jalapeño, developed by Broadcom and primarily designed for data processing and inference.
• Google DeepMind, meanwhile, has relied for years on Alphabet’s TPU chips to power its AI models.
• Meta is also actively developing and deploying its own MTIA accelerators designed for AI workloads.

This trend shows that the integration of hardware and software is becoming a key factor in future success in the AI industry. Initiatives like Anthropic’s could enable the delivery of faster, more efficient and more powerful AI services in the future.