Cognichip, a young startup that wants to use artificial intelligence to design the very chips that power AI systems, has raised 60 million dollars in fresh funding to accelerate its ambitious plan. The Redwood City, California-based company is building what it calls a physics‑informed foundation model for semiconductor design, aiming to dramatically cut both the time and cost involved in bringing new processors to market.
The problem Cognichip is targeting is one of the most stubborn bottlenecks in the technology industry: modern chips are extraordinarily complex and painfully slow to develop. Advanced processors can take three to five years to go from initial concept to mass production, with the design phase alone often stretching on for as long as two years. The latest flagship chips can contain on the order of 100 billion transistors that must be precisely arranged, verified and optimized before any silicon is manufactured.
Cognichip’s pitch is that AI can shoulder a significant part of this burden. The company is developing a deep learning system designed to work alongside human engineers throughout the chip design process, from early architecture to detailed layout. Rather than replacing designers, the AI is intended to handle repetitive, computationally intensive tasks, search huge design spaces and surface solutions that satisfy power, performance and area constraints more efficiently than traditional tools.
“We’re building Artificial Chip Intelligence that essentially can understand, learn and solve chip design problems,” said Stelios Diamantidis, the company’s chief product officer. “We train it with RTL, we train it with post‑synthesis netlist results. We train it with circuits. We even train it, of course, with aspects of specification and validation. But the idea is very simple. ACI understands and performs problem‑solving tasks at a very high speed, and most importantly, at very high parallelism.”
Cognichip refers to its core technology as ACI, short for Artificial Chip Intelligence, a foundation model that is explicitly informed by the physics and engineering constraints of semiconductors. Unlike general-purpose generative AI systems trained primarily on text or images, ACI is trained on design representations and data specific to chips: hardware description languages, circuit topologies, timing reports and layout information.
According to the company, this approach allows the model to reason about trade‑offs in a way that aligns with the realities of chip engineering. ACI is being built to automate and optimize a wide range of tasks, including layout generation, power management strategies and thermal analysis, as well as to identify opportunities to fine‑tune performance and efficiency in existing designs.
Cognichip says its technology can reduce the cost of development projects by more than 75 percent and cut design timelines by more than half. The company argues that this combination of speed‑up and cost reduction could help “democratize” the chip industry by making advanced design capabilities accessible to a broader group of companies that cannot afford multi‑year, multi‑hundred‑million‑dollar projects.
The latest 60 million dollar funding round is a significant vote of confidence in Cognichip’s approach from investors with deep roots in semiconductors. The Series A was led by Seligman Ventures, with participation from Japan‑based SBI Investment and a coalition of other chip‑focused backers; existing investors including Mayfield, Lux Capital, FPV and Candou Ventures also joined in above their pro rata stakes.
The financing brings Cognichip’s total funding to 93 million dollars since its founding in 2024. “Total funding exceeds 93 million dollars as semiconductor leaders rally around Cognichip’s full‑stack ACI technology,” the company said in announcing the round. The new capital will be used to expand its engineering team, deepen work on the ACI platform and scale collaborations with major chipmakers.
As part of the deal, prominent industry figures are also taking seats at the table. The round includes participation from veteran semiconductor executive and investor Lip‑Bu Tan, who will join Cognichip’s board. Umesh Padval, a managing partner at Seligman Ventures, is also joining the board, giving the startup access to decades of experience in chip design, EDA tools and semiconductor company building.
Cognichip emerged from stealth in 2025 with the goal of using generative AI and foundation models to speed up chip development and narrow the gap between rapid software innovation and slower hardware cycles. “Chips are a critical component of the AI industry. But new chips don’t hit the market with the same speed as new AI models and products,” the company said at the time, framing its work as an attempt to bring silicon timelines closer to the pace of AI software.
Industry analysts say the startup is part of a growing wave of companies applying AI to EDA, or electronic design automation, but with a more aggressive, end‑to‑end vision. By training a large model across multiple stages of the design flow and grounding it in physical constraints, Cognichip is trying to create a system that can reason about chips holistically rather than as a series of isolated steps.
“We see this as closing the gap between increased complexity and accelerating software innovation,” one analysis of Cognichip’s approach noted, arguing that physics‑informed foundation models and advanced tooling could make chip development dramatically faster and cheaper, and in doing so open up the field to more players.
The stakes are high because the AI boom has triggered an unprecedented demand for specialized chips, from data‑center GPUs and AI accelerators to custom silicon in consumer devices. Yet the process of designing those chips has not kept pace with the explosion in AI models and applications, creating pressure on the semiconductor industry and contributing to supply constraints.
If Cognichip’s technology delivers on its promises, it could reshape how companies approach the next generation of AI hardware. Faster, cheaper design cycles might allow hardware makers to iterate more often, explore a wider variety of architectures and tailor chips more closely to specific workloads, from large‑language‑model training to low‑power inference at the edge.
Investors clearly see that potential. “This funding round highlights investor confidence in the startup’s potential to disrupt traditional chip design workflows,” one announcement of the deal said, noting the growing interest in AI‑driven design tools among major semiconductor firms. Backers are betting that AI‑assisted design will become a core part of how the industry copes with rising complexity, shrinking process nodes and tightening power and thermal envelopes.
Cognichip will not have the field to itself. Established EDA vendors and large chip companies are also investing heavily in AI‑enhanced tools, and integrating a novel foundation model into conservative, risk‑averse design flows will take time. Chipmakers must be convinced not only that AI‑generated suggestions are correct, but that they can be verified, tested and manufactured at scale without introducing new sources of error.
There are also questions about how far AI can be trusted with design decisions that have billion‑dollar implications if a chip fails in the field. For now, Cognichip is positioning ACI as a co‑pilot that augments, rather than replaces, experienced engineers, and emphasizes that its system is grounded in the physics and constraints that govern real‑world silicon.
With its latest 60 million dollars in fresh capital and a roster of high‑profile backers, Cognichip now has more resources to try to prove that vision in production environments. If it succeeds, the same technology that is transforming software and services could end up redesigning the chips at the foundation of the AI era and doing it faster, cheaper and at a scale the industry has struggled to match by traditional means.
Comments