Credo Technology Group Acquires Israeli Startup DustPhotonics in $1.3 Billion Silicon Photonics Deal

Credo Tech acquires DustPhotonics to lead the silicon photonics market. Learn how light-speed data transfer is reshaping AI infrastructure in 2026.

By: AXL Media

Published: Apr 17, 2026, 8:38 AM EDT

Source: Information for this report was sourced from The Times of Israel

A Massive Multi Billion Dollar Semiconductor Consolidation

The acquisition of Modiin-based startup DustPhotonics by U.S. chip connectivity firm Credo Technology Group represents one of the largest Israeli tech exits of 2026. Under the terms of the agreement, Credo will pay an initial $750 million in cash and 920,000 shares, with the potential for the total value to reach $1.3 billion upon the fulfillment of 3.21 million additional shares linked to financial milestones. This transaction underscores the aggressive push by major hardware providers to secure the underlying components necessary for large-scale AI deployment.

The Strategic Pivot to Optical Connectivity

Silicon photonics technology has emerged as a vital solution for the bottlenecking issues currently facing modern data centers. Unlike traditional copper-based electrical conductors, the technology developed by DustPhotonics utilizes light to transmit data between servers at significantly higher bandwidths. According to DustPhotonics CEO Ronnen Lovinger, this transition to optical processing allows for the rapid transfer of massive datasets required for AI training while simultaneously reducing total power consumption and operational costs.

Market Expansion and the Surge of AI Adoption

The acquisition reflects a broader industry trend as the global silicon photonics market is projected to reach $3.55 billion by the end of 2026. Credo Technology CEO William Brennan noted that as AI adoption accelerates, the underlying infrastructure must move toward a vertically integrated connectivity platform that spans from "chip to cluster." By absorbing DustPhotonics, Credo aims to solve the twin constraints of reliability and power efficiency that limit the scalability of current high-performance computing clusters.