Google Expands Chip Development in Israel with Focus on Edge AI
The company is scaling up its hardware operations in Israel, recruiting senior engineers from Nvidia and other industry leaders to drive breakthrough developments in artificial intelligence.
Google is significantly expanding its chip development operations in Israel by establishing a new group focused on artificial intelligence chips for edge devices, a field known as Edge AI. Concurrently, the company is recruiting a series of executives and experienced engineers from the Israeli high-tech industry to strengthen its hardware capabilities in areas central to the evolution of artificial intelligence.
Leading the new initiative is Guy Kaminitz, former VP of R&D at the chip company Hailo, which was recently sold to Microchip. Alongside him, other experienced hardware engineers from Hailo have joined Google, and the group is currently in the process of recruiting dozens of additional engineers.
The goal is to develop a first generation of AI accelerators capable of performing data processing directly on the device, rather than constantly transferring information to the cloud. This development is aimed at the world of Physical AI, where systems operate within the physical environment and must react quickly and accurately. Relevant fields include robotics, vehicles, and autonomous systems, where response time is critical.
At the same time, Google is strengthening its data communication operations. Ido Bukspan, a former senior executive at Mellanox and Nvidia and CEO of the startup Pliops, has joined the company to lead the development of chips and communication systems for its data centers. Pliops was acquired earlier this year by Astera Labs.
One of the central projects developed in Israel is Axion, a server processor based on ARM architecture, which Google announced in 2024. This is the first processor designed by Google and is currently used in Google Cloud data centers. Simultaneously, teams in Israel are developing hardware technologies and communication chips integrated into Google's AI accelerators in data centers.
This move reflects a broader shift in the AI landscape, where processing is gradually moving from the cloud to edge devices. Local processing reduces dependence on internet connectivity and minimizes latency.
Google, Amazon, and Microsoft are investing significant resources into developing independent chips to reduce reliance on Nvidia. Independent development allows companies to design hardware tailored to their software requirements and maintain better control over their technological infrastructure.
The recruitment of Kaminitz and Bukspan, along with the expansion of engineering teams, highlights the strength of the Israeli hardware industry. For Google, Israel is becoming a significant hub for chip development, ranging from processors and AI accelerators to communication components for data centers.
