Unconventional AI Acquires Israeli Chipmaker NeuReality to Accelerate Development

Unconventional AI, a U.S. hardware startup with a $4.5 billion valuation, announced on Monday it is acquiring Israeli chip startup NeuReality for an undisclosed sum. The cash-and-stock deal is intended to accelerate the company's path to market as its research and development progresses ahead of schedule.

Accelerating Product Roadmap

Unconventional AI CEO Naveen Rao said in an interview with Reuters that the company's progress has been significantly faster than planned. After raising $475 million in December, the firm initially projected a three-year research phase and a five-year product development cycle.

However, Rao noted that by using AI tools to speed its own work, the company has already produced its first test chip and is close to completing a second. "Things have gone a lot faster than anticipated," Rao told Reuters. The acquisition provides an experienced engineering team to help commercialize its technology.

"We need people who can go execute and build products and scale them," Rao added. "We built up the research bones and now we’ve got to build up the production bones."

Deal Details and Strategic Fit

Ad

As part of the acquisition, approximately 40 employees from NeuReality will join Unconventional AI. The company will also establish a presence in Israel. NeuReality, founded in 2019 by veterans of Marvell Technology and Intel, has already developed several of its own advanced AI chips.

Rao, who previously served on NeuReality's board, is a serial entrepreneur who sold chip startup Nervana to Intel for over $400 million and AI software firm MosaicML to Databricks for $1.3 billion.

Context: A New Approach to AI Hardware

Based in Mountain View, California, Unconventional AI launched less than a year ago with the goal of creating AI chips that are 1,000 times more energy efficient than current models. The company is backed by prominent investors including Andreessen Horowitz, Sequoia, and Amazon.com founder Jeff Bezos.

The firm's technology is based on a "dynamical systems" approach, which draws from biological concepts. Instead of using conventional step-by-step logic, this method involves simple computing elements that interact and evolve over time to produce a result, a process intended to drastically reduce power consumption.