Cognichip Raises $60M to Let AI Design the Chips That Power AI

Semiconductor design has never been more important — or more expensive. Advanced chips now take years and billions of dollars to move from concept to silicon, and that bottleneck has become a structural risk for an industry trying to keep pace with AI's demand for compute. Cognichip is betting that AI can fix the problem it helped create. The Toronto-based startup has closed a $60M Series A, bringing total funding to $93 million, to scale its Artificial Chip Intelligence (ACI) platform — a physics-informed AI system purpose-built to automate semiconductor design.

What Cognichip's Platform Actually Does

ACI is not a general-purpose language model applied to chip design. Cognichip's CEO and founder Faraj Aalaei — a twice-over semiconductor IPO executive with AT&T Bell Labs roots — is deliberate about that distinction. "ACI is focused on logic and physics, not poetry or cooking recipes," he has said. The system is trained on the laws that govern semiconductor behavior, enabling it to perform what Aalaei calls "designer-level cognitive reasoning" on silicon problems that historically required human specialists working across months-long cycles.

The platform's headline claims: a 75% reduction in design costs and a 50% cut in development timelines. Cognichip says ACI spans digital, analog, mixed-signal, and foundry environments — the full range of modern chip design disciplines — and is currently engaged with more than 30 semiconductor companies, including many of the top 20 global chipmakers. The company emerged from stealth in May 2025 with a $33M seed round and was named to Fast Company's World's Most Innovative Companies list for 2026.

Who's Backing the Round

The Series A was led by Seligman Ventures, with SBI Investment also participating. Notably, all original seed investors — including Mayfield, Lux Capital, FPV, and Candou Ventures — reinvested above pro rata, a strong signal of inside conviction. But the marquee name is Lip-Bu Tan, Intel's CEO, who is joining Cognichip's board through his vehicle Walden Catalyst. The addition of one of the semiconductor industry's most respected operators as a board member represents a form of validation that capital alone cannot provide.

Seligman's managing partner Umesh Padval also joins the board, alongside scientific advisors Peyman Mohajerin Esfahani and Amin Saberi, director of Stanford's Language, Data, and Reasoning Lab. The founding team draws on alumni from Amazon, Google, Apple, Aquantia, Synopsys, and KLA — a rare combination of hyperscaler and semiconductor-native experience.

Why This Moment

The timing of the raise reflects a growing recognition across the semiconductor industry that the current design methodology cannot scale to meet AI's infrastructure demands. Each new generation of AI hardware requires chips of greater complexity, built on tighter power and area budgets, at shorter intervals. The conventional workflow — teams of engineers using legacy EDA tools in multi-year cycles — was already straining before the generative AI inflection point. Cognichip's thesis is that AI-native design tooling is not a nice-to-have: it is a prerequisite for the industry to keep pace with itself.