Anthropic S-1 Names TSMC and Three HBM Makers as Core Risks

Anthropic has filed to go public, and its Form S-1 names the four companies that hold the keys to its future: TSMC for leading-edge logic and SK hynix, Samsung Electronics and Micron Technology for high-bandwidth memory. The registration statement, filed with the US Securities and Exchange Commission on October 26, 2026, targets a Nasdaq listing under the ticker "ANT." Its risk section reads like a supply chain map, and most of the map sits in East Asia.

What does the filing actually say?

The core warning is straightforward. Anthropic states that its suppliers might not keep pace with rising demand for chips. It also flags a competitive risk: rival AI developers could lock up manufacturing capacity first, which would make shortages worse for everyone else.

HBM is the center of that bottleneck. Standard memory sits a short distance from a processor and feeds it data through a relatively narrow channel. HBM stacks memory chips vertically and places them right next to the processor, creating a much wider pipe. Large AI models move huge amounts of data constantly, and without fast memory even the best accelerator spends much of its time waiting.

Only three companies dominate HBM supply worldwide: SK hynix, Samsung and Micron. Anthropic named all three. For advanced logic, the filing puts TSMC at the top. In plain terms, four suppliers on one edge of the Pacific now underpin the compute plans of one of the most valuable AI companies.

How large are the commitments behind this risk?

The supplier exposure looks sharper next to the money Anthropic has already promised. The filing describes non-cancelable or usage-independent compute agreements worth hundreds of billions of dollars. Research tied to the filing puts those commitments at upwards of $600 billion, with counterparties including Broadcom, Google and Amazon.

The word "non-cancelable" carries most of the weight. Anthropic owes these payments whether or not it uses the compute, and whether or not demand grows as hoped. Fixed costs of that scale make supplier reliability a direct financial question, not an operational detail.

Growth so far has been steep. Revenue went from $386 million in 2024 to approximately $4.585 billion projected for 2025, a nearly twelvefold jump in a single year.

What valuation does the filing point to?

The S-1 does not disclose how many shares Anthropic plans to sell or an expected price range. Private valuations reportedly reached $965 billion early in 2026. Expectations for the IPO valuation range from $1.5 trillion to $2 trillion, according to the report around the filing.

For prospective shareholders, the offering amounts to a two-sided bet. One side is a company growing revenue at a pace few businesses ever reach. The other is a business locked into enormous fixed compute costs while depending on outside suppliers it does not control.

Why does the disclosure matter now?

Anthropic is not alone in this concentration. Any developer of large models needs the same advanced chips and the same high-bandwidth memory. The filing simply puts the problem into a legal document, with Anthropic's name attached. Once a company tells the SEC that its suppliers may not keep up, the issue moves from industry chatter to formal disclosure.

The geography adds another layer. With most key suppliers in East Asia, Anthropic's capacity depends on factories in a small part of the world. The filing also draws fresh attention to the suppliers themselves: TSMC, SK hynix, Samsung and Micron are now formally named as critical inputs to one of the most closely watched IPOs of the AI era, while Broadcom, Google and Amazon sit on the other side of hundreds of billions in committed spending.

Several details will show how the story develops. The share count and price range, when disclosed, will reveal how Anthropic itself values the company. Amended filings may add detail on supplier agreements or reserved capacity. Until then, the supply of advanced logic and HBM remains the variable that decides whether the revenue curve or the cost curve wins.