Anthropic, the Amazon-backed startup behind the popular Claude chatbot, is taking steps toward designing its own artificial intelligence chip. According to a report from The Information, the company has held early-stage discussions with Samsung Electronics about manufacturing the chip and has hired Clive Chan, a former chip engineer at OpenAI, to lead the effort.
The move signals that Anthropic, like several of its AI rivals, is looking to reduce its dependence on Nvidia, whose graphics processing units (GPUs) dominate the market for training and running large language models. By developing custom silicon, Anthropic could potentially lower costs, improve performance, and gain more control over its AI infrastructure.
Why Build a Custom Chip?
Training advanced AI models like Claude requires enormous computing power, and the chips used for that work are in high demand. Nvidia's H100 and newer Blackwell GPUs are the industry standard, but they are expensive and often hard to secure in large quantities. A custom chip, tailored specifically to Anthropic's software, could offer better efficiency and reduce the company's reliance on a single supplier.
Other AI companies have taken similar paths. Google has its Tensor Processing Units (TPUs), Amazon has Trainium and Inferentia chips, and Microsoft has reportedly been developing its own AI chip as well. Even OpenAI, Anthropic's chief competitor, has explored custom chip designs, though it has not yet brought a product to market.
Anthropic's talks with Samsung are still at an early stage, and no deal has been finalized. Samsung is one of the world's largest semiconductor manufacturers, competing with Taiwan Semiconductor Manufacturing Company (TSMC) to produce chips for clients like Nvidia and AMD. If Anthropic moves forward, it would join a growing list of tech companies that design their own chips but rely on foundries like Samsung or TSMC to actually build them.
The Hiring of Clive Chan
The addition of Clive Chan, who previously worked on chip design at OpenAI, gives Anthropic in-house expertise in semiconductor engineering. Chan's background suggests the company is serious about building a dedicated hardware team, a costly and long-term commitment that few AI startups have attempted.
Hiring experienced chip engineers is a competitive field. Companies like Meta, Google, and Apple have been poaching talent from established chipmakers for years. For Anthropic, landing someone with OpenAI experience could accelerate its timeline, though custom chip projects typically take years to yield results.
What It Means for Investors
For everyday investors, Anthropic's chip ambitions are a reminder that the AI boom is increasingly about hardware as much as software. The company is not publicly traded, but its moves have ripple effects. Amazon, which has invested billions in Anthropic, stands to benefit if the startup can lower its computing costs and improve its AI models. Amazon's own chip business, through its AWS cloud unit, could also see synergies if Anthropic's designs align with Amazon's infrastructure.
The broader trend of AI companies building custom chips could eventually challenge Nvidia's dominance, but that shift is likely years away. Nvidia's GPUs remain the gold standard, and its software ecosystem, CUDA, is deeply entrenched. Still, the fact that Anthropic is willing to invest in custom silicon suggests that the current chip supply chain is a bottleneck for AI progress.
Investors should also watch how this affects Samsung. If the South Korean tech giant secures a deal with Anthropic, it would be a win for its foundry business, which has been trying to catch up to TSMC. Samsung's chip division has faced challenges in recent years, including lower yields on advanced manufacturing processes, but landing a high-profile AI client could boost its credibility.
For now, Anthropic's chip plans are exploratory. The company has not confirmed any timeline or budget for the project. But the combination of early talks with a major manufacturer and the hiring of a key engineer signals that Anthropic is thinking beyond software. In the fast-moving AI race, controlling your own hardware could become a critical advantage.
As always, investors should keep an eye on how these developments unfold. The AI chip market is one of the most dynamic areas in tech, and any shift in the competitive landscape could have significant implications for the stocks of Nvidia, AMD, Samsung, and TSMC.
