---
title: "Anthropic Tasks Claude with Designing Its Own Chips: Wall Street Worries Over Synopsys and Cadence"
description: "Anthropic is building an internal team of 15 specialists and using Claude models to design its own chips, triggering Wall Street concerns for EDA vendors Synopsys and Cadence."
date: 2026-11-05T00:11:21.446Z
lang: en
url: https://xab.info/en/posts/anthropic-claude-chip-design-synopsys-cadence
tags: [anthropic, claude, ai-chip-design, semiconductors, synopsys, cadence, eda, openai, asic]
author: "Alexander Buran"
publisher: "XAB.info"
---

# Anthropic Tasks Claude with Designing Its Own Chips: Wall Street Worries Over Synopsys and Cadence

![Artificial intelligence microchip on a motherboard surrounded by processors](https://xab.info/media/2026/11/05/anthropic-claude-proektirovanie-chipov-synopsys-cadence/anthropic-claude-proektirovanie-chipov-synopsys-cadence-1.webp)

## Key Points

- Anthropic has formed a 15-person unit utilizing Claude models to design proprietary chips.
- Synopsys and Cadence shares fell 2.5% and 4.5% amid fears over the EDA sector, which has projected revenues of $16 billion.
- OpenAI previously leveraged AI to develop the Jalapeño accelerator with Broadcom, completing the design in nine months.
- Developing proprietary chips could complicate Anthropic's relationships with investor-suppliers like Amazon, Alphabet, Nvidia, and AMD.
- Anthropic offers research engineers training AI for chip design salaries of up to $850,000 per year.

Anthropic, a key player in the generative artificial intelligence market, has begun actively promoting a new competitive advantage to investors: applying its own AI models during the semiconductor design stage. A dedicated hardware development unit has been formed within the company, currently employing 15 specialists. According to 3DNews, engineers are using Claude models to accelerate the design of their own chips and shorten engineering processes that take significantly longer using traditional development methods employed by Nvidia and AMD.

### AI vs. Multi-Year Roadmaps

Creating internal design models can cut multi-year semiconductor release roadmaps down to just a few months. This fundamentally alters the industry's economics: developers of advanced AI models aim to reduce computing power costs and shorten the lengthy development cycles typical of modern chips. Anthropic is not the first in this direction: when creating the Jalapeño AI accelerator alongside Broadcom, OpenAI already utilized internal AI systems which reportedly allowed chip design to be completed in nine months. Thus, using AI for hardware development is gradually becoming an industry standard among generative AI leaders.

### Market Reaction: Pressure on EDA Vendors

The emergence of such AI-based solutions poses a direct threat to major companies in the electronic design automation and testing software sector. Recently, shares of Synopsys and Cadence dropped by 2.5% and 4.5% respectively. Investors have grown concerned that foundational AI model developers could seriously challenge existing players, with segment revenue projected at $16 billion in the current fiscal year. Meanwhile, Anthropic has not yet announced plans to bring its proprietary semiconductor design tools to the external market, but the mere fact of their internal use is already impacting competitors' stock prices.

### Implementation Barriers and Intellectual Property Protection

The company has already approached electronic design automation developers requesting access to proprietary datasets to improve its own models. However, the widespread adoption of such tools by third-party companies may face resistance: chip developers fiercely protect their design methodologies and often prefer isolated infrastructures or open-source solutions over external AI models to eliminate the risk of intellectual property leaks. This remains one of the key constraints on scaling the AI-driven approach to silicon design.

### Talent Stakes and Multi-Chip Approach

According to Playground, Anthropic is building an internal silicon development team and intends to pursue a multi-chip approach: the idea is not only to create its own processor but to leverage AI for the hardware development process itself. Salaries within the new unit vary noticeably by role: the company is willing to pay research engineers training AI models for chip design between $500,000 and $850,000 per year, while specialists directly developing the company's first proprietary ASIC are offered a range of $320,000 to $485,000. Requirements for both categories largely overlap—full-cycle ASIC and FPGA development experience, RTL-to-tape-out, UVM, formal verification, physical design, PPA, DFT, and EDA tool proficiency.

### Potential Conflict with Financial Sponsors

Developing custom chips could spark friction between Anthropic and its main financial backers, who have invested a combined total of around $90 billion in the company while simultaneously acting as its primary infrastructure and cloud solution providers. Relationships with partners such as Amazon, Alphabet, Nvidia, and AMD could become strained, even though internal models can drastically accelerate semiconductor rollouts. Consequently, Anthropic's strategy lies at the intersection of technological ambition and commercial obligations to investors, making it one of the most closely watched industry cases for 2026.

## Fact-Check & Sources

- [Anthropic поручила Claude проектировать чипы — на Уолл-стрит испугались за Synopsys и Cadence window-new](https://3dnews.ru/1150832/anthropic-poruchila-claude-proektirovat-chipi-na-uollstrit-ispugalis-za-synopsys-i-cadence) - Primary source: 15-person unit, Claude usage, Synopsys (2.5%) and Cadence (4.5%) stock drops, EDA revenue of $16B, investments around $90B, OpenAI/Broadcom Jalapeño 9-month example.
- [Anthropic платит до 850 тысяч долларов инженерам, обучающим ИИ проектировать чипы - создателям ASIC ...](https://www.playground.ru/misc/news/anthropic_platit_do_850_tysyach_dollarov_inzheneram_obuchayuschim_ii_proektirovat_chipy_sozdatelyam_asic_zametno_menshe-1865122) - Additional context: salaries of $500–850k for research engineers and $320–485k for ASIC developers, multi-chip approach, confirmation of custom accelerator plans.

## FAQ

### How many specialists work in Anthropic's hardware development unit?

According to 3DNews, the unit employs 15 specialists who use Claude models to design proprietary chips.

### How did the market react to news regarding Anthropic's chips?

Synopsys and Cadence shares dropped by 2.5% and 4.5% respectively amid investor concerns regarding the EDA sector.

### Did anyone use AI for chip design before Anthropic?

Yes, OpenAI used internal AI systems when developing the Jalapeño accelerator alongside Broadcom, completing the design in nine months.

### Could chip development affect Anthropic's relationships with investors?

Yes, it could potentially complicate ties with Amazon, Alphabet, Nvidia, and AMD, which have invested around $90 billion while serving as infrastructure and cloud providers.