Cognichip Raises $60M for AI-Designed AI Chips
▼ Summary
– Cognichip is developing an AI model to assist engineers in designing computer chips, aiming to address the industry’s slow, expensive, and complex design process.
– The company claims its technology can reduce chip development costs by over 75% and cut the timeline by more than half.
– Cognichip recently raised $60 million in funding, bringing its total to $93 million, with Intel CEO Lip-Bu Tan joining its board.
– Its AI model is trained on proprietary and synthetic chip design data, as the required data is closely guarded and not openly available like software code.
– The firm faces competition from established companies like Synopsys and well-funded startups such as ChipAgentsAI and Ricursive.
Artificial intelligence has been supercharged by the world’s most sophisticated silicon. Now, a new venture is turning that relationship on its head, using AI to fundamentally redesign how we create the chips themselves. Cognichip is developing a deep learning model to act as a co-pilot for semiconductor engineers, tackling an industry-wide challenge that has persisted for generations: the extreme complexity, staggering cost, and painfully slow pace of chip design. Moving a state-of-the-art processor from concept to factory can consume three to five years, with the initial design phase alone stretching up to two years. For perspective, Nvidia’s latest Blackwell GPU architecture packs 104 billion transistors, a monumental task to organize manually.
Cognichip’s founder and CEO, Faraj Aalaei, argues that by the time a traditional design cycle concludes, market demands can shift, rendering a massive investment obsolete. His mission is to import the AI-powered productivity tools common in software engineering into the semiconductor realm. “These systems have now become intelligent enough that by just guiding them and telling them what the result is that you want, it can actually produce beautiful code,” Aalaei stated. The company claims its technology can slash development costs by over 75% and reduce timelines by more than half.
Emerging from stealth operations last year, Cognichip recently announced a $60 million funding round. The investment was led by Seligman Ventures and included significant participation from Intel CEO Lip-Bu Tan through his firm, Walden Catalyst Ventures. Tan will join Cognichip’s board alongside Umesh Padval, a managing partner at Seligman. This latest capital infusion brings the company’s total funding to $93 million since its 2024 founding.
A key hurdle for Cognichip has been acquiring the specialized data needed to train its models. Unlike the software world, where developers freely share code, chip design IP is fiercely guarded proprietary information. This lack of an open-source repository forced Cognichip to build its own datasets, utilizing synthetic data and licensing information from partners. The firm also created secure procedures allowing chipmakers to train models on their confidential data without risk of exposure. Where proprietary data was inaccessible, the company turned to open-source alternatives. In a 2025 demonstration, electrical engineering students at San Jose State University used Cognichip’s model in a hackathon to design CPUs based on the open RISC-V architecture.
The competitive landscape is intense. Cognichip faces established electronic design automation giants like Synopsys and Cadence, as well as heavily funded newcomers. These include ChipAgentsAI, which secured a $74 million extended Series A in February, and Ricursive, which raised a $300 million Series A in January. Despite its progress, Cognichip has not yet publicly showcased a chip fully designed with its system and has not named the collaborators it says it began working with last September.
Commenting on the market dynamics, investor Umesh Padval noted the unprecedented flow of capital into AI infrastructure, calling it the largest surge he has witnessed in four decades. “If it’s a super cycle for semiconductors and hardware, it’s a super cycle for companies like [Cognichip],” he observed. The race to accelerate chip creation is clearly heating up, fueled by the very technology these chips are built to run.
(Source: TechCrunch)
