By Connie · This article contains affiliate links. We may earn a commission at no extra cost to you if you sign up through our links.
Cognichip Raises $60M So AI Can Design the Chips That Run AI
By Connie · April 1, 2026 · 7 min read
Cognichip raised $60M on April 1, 2026 — led by Seligman Ventures with Intel CEO Lip-Bu Tan joining the board — to build ACI® (Artificial Chip Intelligence), an AI platform that designs semiconductor chips. ACI® cuts chip development costs by 75% and design timelines by 50%+. Total raised: $93M. This is the clearest sign yet that AI is now accelerating its own hardware infrastructure.
What Just Happened: Cognichip Closes $60M Round
Cognichip, the AI startup building what it calls Artificial Chip Intelligence (ACI®), announced a $60 million funding round this morning. The round was led by Seligman Ventures, with a standout strategic participant: Lip-Bu Tan, the CEO of Intel, who invested through his venture firm Walden Catalyst Ventures and will join Cognichip's board of directors.
The company was founded in 2024 by Faraj Aalaei, an experienced semiconductor executive, alongside CPO Stelios Diamantidis — a 25-year veteran of EDA (Electronic Design Automation) and AI-driven chip design. Cognichip launched from stealth in May 2025 with a $33 million seed round from Mayfield, Lux Capital, FPV, and Candou Ventures. Today's raise brings total funding to $93 million.
Cognichip was also named to Fast Company's World's Most Innovative Companies of 2026 in March, signaling growing industry recognition before today's funding announcement.
What ACI® Does: AI That Thinks Like a Chip Designer
Traditional chip design is one of the most complex engineering disciplines on earth. Modern processors — the GPUs and custom ASICs that power AI systems at data centers — contain billions of transistors and require teams of hundreds of engineers working for 4-7 years to bring a single chip from concept to silicon. The process is linear, failure-prone, and enormously expensive. A single chip tape-out can cost $50-100M.
Cognichip's ACI® platform is trained specifically on chip design data — not general code or generic problem-solving, but the deep domain knowledge of hardware engineers: placement, routing, power analysis, timing closure, verification. The model can understand chip design problems, generate solutions, evaluate tradeoffs, and iterate — mimicking what a senior chip designer does, at machine speed.
Existing EDA (Electronic Design Automation) tools from companies like Synopsys and Cadence help engineers design chips but require constant human decision-making. ACI® targets a different model: autonomous AI that handles entire design sub-problems end-to-end, with engineers reviewing outputs rather than driving every step.
While Cognichip uses AI to design chips, Happycapy uses AI to accelerate knowledge work — writing, research, coding, and analysis. Start free and see what AI-augmented productivity feels like.
Try Happycapy FreeThe Recursive Loop: AI Accelerating AI
The deeper significance of Cognichip's work is what it implies for the trajectory of AI development itself. Today, AI capability is constrained by hardware — the Nvidia GPUs, Google TPUs, and custom ASICs that run AI models. Those chips are slow and expensive to design, creating a natural bottleneck on how quickly the AI industry can scale.
If AI can design better chips faster and cheaper, that bottleneck breaks. Better chips enable more capable AI models. More capable AI models design even better chips. The loop compounds. This is what researchers sometimes call the "recursive self-improvement" pathway in hardware — and Cognichip is one of the first well-funded companies trying to unlock it commercially.
Why Intel's CEO Joining the Board Is the Real Signal
Venture rounds happen every day. A sitting CEO of one of the world's largest semiconductor companies joining the board of a 2-year-old AI chip startup is something else entirely.
Lip-Bu Tan took over as Intel CEO in March 2024 with a mandate to rebuild Intel's foundry business and reclaim its manufacturing leadership. Intel's core challenge is designing next-generation chips faster than TSMC and Samsung-backed competitors while cutting costs. If ACI® can compress Intel's internal design cycles and reduce tape-out failure rates, the strategic value is enormous — orders of magnitude larger than the $60M investment.
Tan's participation signals this is not a passive financial bet. It suggests Intel sees ACI® as potentially transformative for its own design operations, and wants a board seat to influence the technology's direction.
Who else is in this space
Cognichip's bet is that a purpose-built AI system — trained from scratch on chip design data, not adapted from a general LLM — will outperform the AI bolt-ons that incumbents like Synopsys and Cadence have added to their existing tools.
What This Means for the AI Hardware Race in 2026
The AI hardware supply chain is the choke point of the entire AI industry. Every major AI lab — OpenAI, Google DeepMind, Anthropic, Meta, xAI — is constrained by access to compute. Nvidia's GPUs remain the dominant option, but demand exceeds supply, prices are high, and the next generation of chips (Nvidia Vera Rubin, Google TPU v6, custom OpenAI silicon) won't reach at scale until 2027 or later.
Startups like Cognichip are betting that compressing design cycles is the fastest path to changing that dynamic. If you can design a new chip in 2.5 years instead of 5, you can iterate on architectures twice as fast. If each design attempt costs $15M instead of $60M, smaller players can afford to tape out and experiment.
The broader investment trend supports this view. AI infrastructure companies — data centers, networking, power, and now chip design — are attracting capital at record rates in early 2026. Nvidia's recent $2B investment in Marvell Technology (March 31) and Cognichip's $60M today are two data points in the same larger shift: the AI industry is investing heavily in the foundational layers that will determine which companies can scale AI fastest.
Happycapy runs on the best AI available today — Claude, GPT-5, and more — so you get maximum capability right now for writing, research, and analysis at $17/mo.
Start Free with HappycapyFrequently Asked Questions
Cognichip is an AI startup founded in 2024 by CEO Faraj Aalaei that uses AI to automate semiconductor chip design. Its platform, ACI® (Artificial Chip Intelligence), mimics human chip designer cognitive abilities — understanding, learning, and solving design problems across a broad range of chip types.
Cognichip has raised $93 million in total — a $33 million seed round when it launched from stealth in May 2025, followed by a $60 million round announced April 1, 2026, led by Seligman Ventures with participation from Intel CEO Lip-Bu Tan through his Walden Catalyst Ventures firm.
Cognichip claims ACI® reduces chip development costs by more than 75% and cuts design timelines by more than half compared to traditional design processes.
AI chips take 4-7 years and billions of dollars to design. If AI can design better chips faster and cheaper, it creates a recursive feedback loop: better chips enable more capable AI, which designs even better chips. Cognichip is targeting that loop directly.
- Tim Fernholz, TechCrunch — "Cognichip wants AI to design the chips that power AI, and just raised $60M to try" (April 1, 2026)
- Cognichip — Official company announcements and ACI® platform documentation (2025-2026)
- Fast Company — World's Most Innovative Companies of 2026 (March 24, 2026)
- SemiEngineering — "Cognichip: Using AI To Speed Complex Chip Design" (August 2025)
- Happycapy AI — AI workspace for knowledge workers
Get the best AI tools tips — weekly
Honest reviews, tutorials, and Happycapy tips. No spam.