The most sophisticated silicon chips have propelled artificial intelligence forward. Now, the question is whether AI can reciprocate. Cognichip is developing a deep learning model intended to collaborate with engineers during the design of new computer chips. It aims to tackle a longstanding industry challenge: chip design is incredibly complex, prohibitively expensive, and time-consuming. Advanced chips require three to five years to progress from initial concept to mass production, with the design phase alone potentially lasting up to two years before physical layout even starts. To illustrate the scale, consider that Nvidia's latest Blackwell GPUs contain 104 billion transistors—an immense number to arrange correctly.

Cognichip's CEO and founder, Faraj Aalaei, notes that during the lengthy process of creating a new chip, market conditions can shift, potentially rendering the entire investment obsolete. His objective is to introduce AI tools similar to those that have accelerated software engineering into the semiconductor design arena. He claims the company's technology can slash chip development costs by over 75% and reduce the timeline by more than half.

The company came out of stealth last year and announced on Wednesday that it secured $60 million in new funding. The round was led by Seligman Ventures, with significant participation from Intel CEO Lip-Bu Tan, who invested through his venture firm Walden Catalyst Ventures and will join Cognichip's board. Umesh Padval, a managing partner at Seligman, will also take a board seat. In total, Cognichip has raised $93 million since its founding in 2024.

EMBED_PLACEHOLDER_0

However, Cognichip cannot yet showcase a new chip designed using its system and has not revealed the names of any customers it says it has been working with since September. The company asserts its edge comes from using its own model, trained specifically on chip design data, rather than adapting a general-purpose large language model. Acquiring the necessary domain-specific training data was a significant hurdle. Unlike software developers, who openly share vast code repositories, chip designers protect their intellectual property closely. This means the open-source datasets typically used to train AI coding assistants are largely inaccessible in this field.

To overcome this, Cognichip has created its own datasets, including synthetic data, and licensed data from partners. It has also established secure procedures allowing chipmakers to train Cognichip's models on their proprietary data without exposing it. Where proprietary data is unavailable, the company has turned to open-source alternatives.

In a demonstration last year, Cognichip invited electrical engineering students at San Jose State University to test the model in a hackathon. The student teams successfully used it to design CPUs based on the open-source RISC-V chip architecture—a freely available design that anyone can build upon.

Cognichip faces competition from established industry leaders like Synopsys and Cadence Design Systems, as well as a group of well-funded startups. These include Alpha Design AI, which raised a $21 million Series A in October 2025, and ChipAgentsAI, which closed a $74 million extended Series A in February.

Padval observed that the current surge of investment into AI infrastructure is the largest he has witnessed in his 40-year investing career. "If this is a super cycle for semiconductors and hardware, it's a super cycle for companies like [Cognichip]," he stated.