From the latest consumer electronics to new applications of artificial intelligence (AI), a single challenge often hinders competitors in the race to launch advanced technologies. Translating the microchip theory at the heart of these innovations into ready-to-install manufacturing components can take up to four years. This challenge only grows more daunting as chips continue to grow in complexity.

"We urgently need to see an acceleration of chip development cycles," said Harald Kröll, co-founder and CEO of the Swiss startup Chipmind. "Technological progress is being hampered by the complex, time-consuming, and costly chip development process it relies on."

Chipmind, which announced today the closing of a $2.5 million pre-seed funding round, believes it can help. The Zurich-based company has developed a set of artificial intelligence agents that chipmakers can use to design and test new chips. Its goal is to automate the menial and routine tasks currently performed manually by engineers; Chipmind estimates that this type of work accounts for about 40% of development work. Does this mean the engineers involved could lose their jobs?

The company is the latest European tech startup to emerge from ETH Zurich, where Kröll met co-founder and CTO Sandro Belfanti a decade ago. The two subsequently collaborated in the fields of AI and chip design, experiencing firsthand the frustrations that slow development cycles.

"Anyone who has worked in chip development knows that the work is repetitive, time-consuming, and requires precision, but not necessarily creativity," Belfanti says. "I often wish there were a solution that would magically handle these mundane tasks so I could focus on solving real engineering challenges."

Is Chipmind that magical solution? Kröll and Belfanti are working with a handful of chip manufacturers, primarily in Europe, where development cycles are particularly slow, on proof-of-concept projects to demonstrate that the technology can work. Crucially, Chipmind believes that general-purpose AI tools won't succeed in this market—instead, it says, intelligent agents need to be able to learn from each manufacturer's systems and environments, customizing their development approach to each task.

In this sense, the challenge is slightly different from software development, as today AI agents typically handle the bulk of the work involved in developing new code. But Kröll sees Chipmind's work as similar to what's happening in the software market. "We've seen AI agents being used to significantly speed up the development of new software, but nothing like that has happened in hardware yet."

In fact, it's a surprisingly underdeveloped area of the market. Two of the most prominent chip design experts—American companies Cadence Design Systems and Synopsys—have both publicly committed to developing AI-based chip development systems, posing a significant competitive threat. Startups are also actively participating, with California-based ChipAgents raising a seed round of just over $3 million late last year. In contrast, in Europe—where manufacturers may be most in need of accelerating chip development—innovation has been constrained.

Chipmind believes its technology can shave as much as a year off the four-year development cycle for new chips, a transformative acceleration. "Our goal is for our agents to help manufacturers solve real-world problems they face every day," Kröll added.

The company hopes to achieve full commercial production in the second half of next year. This will require further R&D investment, and today's funding is intended to support the expansion of the engineering team, accelerated product development, and deeper collaborations with key industry players.

The round was led by Swiss seed fund Founderful. "In a world where artificial intelligence is becoming increasingly ubiquitous, Chipmind stands out by offering a refreshing solution to a problem that Harald and Sandro have been working on for 20 years," said Edouard Treccani, Head of Founderful.

Transforming Semiconductor Design with Agent AI

Meanwhile, Santa Barbara-based ChipAgents.ai announced today that it has raised $21 million in early-stage funding to advance its agent-based AI platform for chip design and verification.

Bessemer Venture Partners led the Series A round, with strategic support from semiconductor giants such as Micron Technology, MediaTek, and Ericsson. Existing investors ScOp Venture Capital and Amino Capital also participated. This round brings the company's total raised to date to $24 million.

Professor William Wang, founder and CEO of ChipAgents, told SiliconANGLE that formal chip design began in the 1980s, when electrical engineers were still working on paper, developing gate architectures. But with the advent of AI chips and new system-on-chip (SoC) designs, the number of logic gates has reached staggering levels of complexity.

"Chips have become so complex, from millions of logic gates to billions, tens of billions, and even trillions," Wang said. "So when you have to push the design complexity to that level, no one can understand the code."

Today, engineers use advanced tools to visualize and construct chip architectures, but traditional software struggles. To address this challenge, ChipAgents has developed an agent AI platform that automates routine design and verification tasks, allowing engineers to focus on innovation rather than the tedious work of writing line-by-line code.

ChipAgents operates at the front end of the chip design and verification process, before chip manufacturing begins. Engineers use it from the earliest stages of defining chip specifications, generating hundreds of pages of PDF files and datasheets.

Wang noted that companies like Cadence Design Systems Inc. and Synopsys Inc. were early pioneers in simulation-based chip design and electronic design automation (EDA), and ChipAgents now aims to enhance these processes with agent AI-powered tools.

Artificial Intelligence (AI)

The platform leverages artificial intelligence (AI) technology to help engineers identify inconsistencies or defects by cross-checking specifications across multiple documents, generating register transfer level (RTL) code, and automatically generating documentation. RTL is a hardware design abstraction that describes how digital circuits transfer data and how logic executes at the hardware level. It is typically written in hardware description languages such as Verilog or VHDL.

Wang stated that ChipAgents' generative AI capabilities allow architecture, design, and verification teams to streamline their workflows and reduce time to market.

A primary use case for the company's platform is verification. Unlike software engineering, where developers can readily fix defects if they occur, hardware verification requires chips to be near-perfect before production, as any error could cost millions of dollars.

"In the industry, people spend more time verifying the functional correctness of chips because it's easy to write code, but how do you know it's actually correct?" Wang said.

He added that each design engineer typically has two to three verification engineers. ChipAgents accelerates verification engineers' work by automatically generating testbenches, rules, and assertions—tasks that once took weeks can now be completed in minutes. This helps teams quickly confirm that chip implementations meet their specifications.

While Wang declined to name specific customers, he stated that many of the top 20 semiconductor companies in the world use ChipAgents. Founded in 2024, the company has experienced explosive growth, with usage surging 60-fold in the first half of 2025, though the exact base has not been disclosed.

"We're seeing significant usage and new use cases every day," Wang said. "I believe we're witnessing a shift in the semiconductor industry toward agent-based AI solutions for design verification."

With the new funding, ChipAgents plans to further strengthen its R&D efforts and focus on providing expanded customer support. This will help improve on-the-ground support for semiconductor customers, who often manage multi-million dollar chip projects.

A key initiative, Wang explained, is the establishment of a new headquarters in Santa Clara, bringing the company closer to the heart of Silicon Valley and expected to be operational within weeks. The company's current base in Santa Barbara will remain its primary R&D center.

"ChipAgents' agent-based AI chip design approach integrates seamlessly into the overall chip design workflow," said Lance Co Ting Keh, venture partner at Bessemer Venture Partners. "We believe integrating disparate EDA tools, from specification extraction to waveform analysis, is the right approach to address this complex, multi-step process."

Source: Content compiled from Forbes

Reference Link

https://www.forbes.com/sites/davidprosser/2025/10/21/inside-the-race-to-accelerate-chip-development-with-ai

View more at EASELINK