Ex-Apple, Amazon engineer launches AI chip startup in his 50s

At 55 years old, Stephen Huang did what many would consider unthinkable: he walked away from a stable, high-paying career in Silicon Valley to launch an AI chip startup from scratch. The catalyst? ChatGPT’s explosive debut in late 2022, which convinced him the moment for his long-held vision had finally arrived.

“I felt the market had arrived,” Huang said.

His resume reads like a who’s who of hardware engineering: GPUs at MediaTek, Apple’s Face ID technology, and a stint on Amazon’s AI chip team. But instead of coasting toward retirement, Huang founded Tranxform AI in 2024. Based in Taiwan, the company is developing power-efficient processors designed to run AI models outside massive data centers. The startup now employs roughly 40 people and expects its first chip to be ready next year.

Huang is part of a growing wave of entrepreneurs capitalizing on the AI boom. But unlike many founders building companies around large language models, his expertise is rooted in decades of chip design. And he never saw his age as a liability.

“Morris Chang started TSMC in his 50s,” Huang noted, referencing the founder of the world’s largest contract chipmaker. He believes hardware startups often benefit from experience in ways software companies do not. Building a system-on-a-chip (SoC) , the integrated processor powering everything from smartphones to AI systems , requires balancing countless tradeoffs across hardware and software. “To build a good SoC, you need experience,” he said. “Otherwise, you would not know how to balance different operations.”

The leap was far from obvious. Before Tranxform, Huang enjoyed a comfortable life in the U. S. Launching the company meant taking a significant risk and spending most of his time in Taiwan , a decision his family initially struggled to accept. As the startup hit key milestones, their attitude shifted. “Today, they are proud of what we have accomplished,” he said.

Timing also played a role. With both sons now adults , one working in tech, the other a recent university graduate , Huang found the freedom to dedicate himself fully. “Having them become independent has made it easier for me to dedicate the time and energy needed to build Tranxform,” he said.

Huang is betting that demand for specialized AI hardware will continue to grow as companies seek faster, more energy-efficient ways to run increasingly complex models. His optimism is backed by market trends. According to PitchBook, venture funding for AI and machine-learning chip startups surged over 70% to $16.2 billion in 2025 from the year before, though deal counts fell to 232 from 266 as investors concentrated capital on fewer potential breakout companies. As of June 22 this year, funding stood at $9.9 billion across 87 deals.

Returning home

While the AI boom drew Huang toward entrepreneurship, Silicon Valley’s intensifying talent wars made him rethink where to build. Years in the Valley taught him how hard it was for startups to compete with giants like Google, Apple, and Nvidia for engineering talent. “We kept training people, and they got poached,” he said.

He built Tranxform in the Taiwanese chip hub of Hsinchu, where he believed he could assemble a more stable team. One key recruit was his college classmate Way-Shing Lee, who joined as chief technology officer after retiring from Qualcomm.

Now, startup life has replaced stability with fundraising, recruiting, customer meetings, and constant problem-solving. “Starting a company is very hard,” Huang said. “You have to find business partners. You have to sell your story. You have to find funding.”

Tranxform is preparing for its next fundraising round, though Huang declined to share details. The company remains in early licensing stages and generates little revenue. Still, he believes the biggest opportunities in AI are ahead. The industry, he said, is “probably just getting started.”