The U.S. government will award $300 million to GlobalFoundries to accelerate development of silicon photonics technology, the company said Wednesday. The funding is part of the Trump administration's push to steer CHIPS Act research dollars toward critical semiconductor technologies, as Washington works to build domestic supply chains for optical components now concentrated in Asia and Israel.
What the funding covers
The $300 million from the Department of Commerce will support research on new materials and techniques for producing and packaging optical components alongside AI processors. The work targets co-packaged optics, which integrates silicon photonics directly with compute chips to improve data-transfer speeds and energy efficiency. The award follows a $1.5 billion grant to GlobalFoundries in 2024 for factory expansions in Malta, New York, and Burlington, Vermont. The administration has also committed $150 million for semiconductor manufacturing equipment and $2 billion for quantum computing.
How silicon photonics works
Instead of electrical signals, silicon photonics uses light to move data between chips. This approach enables higher bandwidth and lower power consumption, which matters for AI workloads where massive data flows between accelerators. GlobalFoundries is targeting transfer speeds of 400 gigabits per second and up to five times the energy efficiency of current implementations.
GlobalFoundries' end-to-end ambition
Most silicon photonics and advanced optical packaging today happens outside the U.S., at companies like TSMC and Tower Semiconductor. GlobalFoundries aims to offer a single-source solution. "To have one company with end-to-end service, starting with a bare silicon wafer and being able to ship a known good and tested optical module in the United States, we think that's a unique capability for us," said Gregg Bartlett, GlobalFoundries' chief technology officer, in an interview with Reuters.
Why this matters for IT and development professionals
For infrastructure teams building AI data centers, optical interconnects directly affect throughput and power budgets. As models grow, the bottleneck shifts from compute to data movement. Domestic production of silicon photonics could shorten supply chains and alter the cost structure of high-speed networking hardware. Professionals looking to stay ahead of hardware trends can find relevant training through AI for IT & Development resources.
Your membership also unlocks:
