optical AI

California-based startup Volantis Semiconductor has raised $88 million to address a key bottleneck in AI computing: powerful chips often sit idle waiting for data to arrive. The company is developing optical connections that move data using light instead of electrical signals, promising faster data transfer between chips. The funding round is drawing attention as demand for AI compute infrastructure keeps climbing.

GlobalFoundries has received an additional $300 million in funding under the US CHIPS Act from the Trump administration, strengthening its position in American semiconductor manufacturing. The news comes as Nvidia CEO Jensen Huang called silicon photonics 'essential' to the AI future, highlighting growing demand for advanced optical chip technologies that companies like GlobalFoundries are positioned to supply.

A new market forecast predicts the optical input/output market for AI chips will grow to $14 billion by 2035. The projection reflects the semiconductor industry's shift toward optical interconnects as electrical links struggle to keep up with the bandwidth and power demands of large-scale AI datacentres, signalling a major opportunity for photonics suppliers.

Optical interconnect startup CScale has come out of stealth, revealing investment from Nvidia and Intel. The company is developing optical interconnect technology aimed at improving data transmission between chips, a key area as AI data centers demand faster, more energy-efficient networking. Backing from both major chipmakers signals strong industry interest in next-generation optical connectivity for large-scale computing.

Researchers have developed an AI system powered by light rather than conventional electronics that can identify deepfake images with nearly 98% accuracy. The optical computing approach could offer a faster, more energy-efficient way to fight AI-generated disinformation as fake media becomes harder to distinguish from real content.

A startup specializing in optical interconnect technology for AI scale-up systems has secured backing from both Nvidia and Intel, according to industry outlet DigiTimes. The investment signals growing interest in optical solutions to handle the massive data bandwidth demands of AI computing clusters, as chipmakers look beyond traditional copper connections for faster, more energy-efficient links between processors.

Researchers with the UNLOC project at Pennsylvania State University report developing exceptional electro-optic thin films, materials that convert electrical signals into optical ones with high efficiency. Such films are considered key building blocks for faster optical modulators, with potential uses in quantum information systems and AI hardware where data transfer speed and energy efficiency are critical.

Marvell Technology is drawing attention after showcasing its 2nm optical link technology and securing manufacturing capacity for AI-related demand. The chipmaker is positioning itself as a key supplier of high-speed interconnects for data centers, where optical links are critical to moving data between processors in AI infrastructure. Investors and industry watchers are tracking the announcement for what it signals about Marvell's role in the AI hardware buildout.

AT&T and Corning announced a partnership to expand and engineer fiber optic networks, in what the companies call an effort to 'engineer the connected world.' The collaboration ties AT&T's broadband expansion goals to Corning's fiber and cable manufacturing, as demand for high-speed connectivity keeps growing with cloud services and AI workloads. Details on investment size and deployment timelines were not included in the initial announcement.

CScale, a startup developing optical AI chips, has come out of stealth and announced $145 million in funding. The company joins a growing wave of hardware firms betting that light-based computing can cut the cost and energy demands of running AI models. Details on its investors and product roadmap remain limited as it begins operating publicly.