STMicroelectronics (ST) has quietly acquired Toronto-based AI startup Deeplite, according to an announcement by Deeplite CEO Nick Romano on social media. While neither party disclosed the financial terms, Deeplite confirmed that it has become a wholly owned subsidiary of ST.

"We are excited to join STMicroelectronics to continue advancing edge AI solutions," Deeplite stated. "Our expertise in model optimization, quantization, and compression will further enable ST to deliver efficient, high-performance AI at the edge."

The acquisition highlights ST's focus on Deeplite's unique technology. While most AI applications today, such as chatbots and object detection systems, rely on cloud-based training and deployment, Deeplite's software enables AI to run efficiently on-device, such as on smartphones and robots. With a client base largely made up of semiconductor companies, Deeplite's offerings align closely with ST's core business of chip design and manufacturing.

By combining Deeplite's advanced edge AI software with its own cutting-edge microcontrollers (MCUs) and neural processing units (NPUs), ST plans to build one of the most advanced edge AI platforms in the world.

Deeplite was founded in 2017 by Ehsan Saboori and Davis Sawyer at the TandemLaunch incubator in Montreal. Nick Romano joined in 2019 as the third co-founder and CEO, spinning Deeplite off into an independent company. The startup, headquartered in Montreal and Toronto, had raised $6.47 million from Canadian investors including BDC Capital, Desjardins Capital, and Somel Ventures, according to PitchBook.

Deeplite specializes in making deep neural networks faster, smaller, and more energy-efficient, facilitating AI deployment in edge devices such as security cameras, sensors, drones, smartphones, and vehicles. The company has emphasized the growing challenges of deploying AI models on edge devices, including technical complexity, compatibility issues, limited resources, and data quality concerns — all of which can impact the sales and adoption of AI processors.

Deeplite has also developed an "Edge AI Maturity Matrix" to help semiconductor companies assess their readiness to support AI deployments. It emphasizes that successful edge AI strategies require not just powerful processors, but also robust model libraries, easy-to-use deployment tools, and pre-optimized AI models to accelerate time-to-market, lower costs, and drive processor sales.

The acquisition comes as ST is reshaping its global manufacturing footprint. On April 10, ST outlined plans to increase investment in AI and automation technologies, aligning perfectly with its move to absorb Deeplite's capabilities.

Meanwhile, ST reported first-quarter 2025 revenue of approximately $2.52 billion, down 27.3% year-on-year and 24.2% sequentially. Its gross margin fell to 33.4%, while net income reached $56 million. Despite the revenue dip, ST noted improvements in its book-to-bill ratio, with automotive and industrial segments outperforming the average.

The acquisition of Deeplite marks a strategic step for ST to make AI more accessible and affordable across edge applications, further positioning the company to capitalize on one of the semiconductor industry's biggest growth opportunities.