Credo acquires DustPhotonics for up to $1.3B to expand Silicon Photonics and AI connectivity

Credo Acquisition Transforms AI Infrastructure Landscape

Credo Technology Group Holding Ltd. has entered a definitive agreement to acquire DustPhotonics Ltd., an Israeli Silicon Photonics startup, for up to $1.3 billion in a deal structured to capitalize on surging demand for AI infrastructure 2

2

. Under the terms, Credo will pay $750 million in cash and issue approximately $123 million in shares at closing, with up to $430 million in additional shares contingent on DustPhotonics hitting certain financial milestones 1

1

2

2

. Shares in Credo rose around 13% following the announcement, reflecting investor confidence in the strategic move 2

2

.

Advanced Silicon Photonics Technology Strengthens Optical Connectivity Solutions

Founded in 2017, DustPhotonics brings a portfolio of Silicon Photonics Photonic Integrated Circuit technology spanning 400G, 800G optical transceivers and 1.6T speeds, with a roadmap extending to 3.2T 2

2

. The company's PIC technology integrates multiple optical functions onto a single chip rather than assembling them from discrete components, reducing complexity, improving manufacturing yields and lowering costs as speeds scale beyond 800G 1

1

. DustPhotonics operates as a fabless semiconductor company, employing approximately 70 engineers with expertise in photonic integration 2

2

. Its technology is already deployed in optical transceivers at leading hyperscale AI clusters, with further applications in near-port optics and co-packaged optics under development 1

1

2

2

.

Vertical Integration Expands AI Connectivity Capabilities

By acquiring DustPhotonics, Credo is positioning itself as a vertically integrated connectivity powerhouse spanning SerDes, DSP, SiPho PICs and system-level integration 1

1

. This comprehensive stack enables Credo to deliver electrical and optical interconnects for both AI scale-out networks and scale-up architectures 2

2

. "Combining forces with DustPhotonics marks a defining step in Credo's strategy to lead across the full spectrum of AI connectivity," said William Brennan, chairman, president and chief executive officer of Credo Technology 2

2

. By bringing Silicon Photonics in-house, Credo reduces supplier dependence, accelerates innovation and enables end-to-end optimization of the signal chain for AI workloads 1

1

.

Financial Projections and AI Data Center Capabilities

Through this Credo acquisition, the company expects its combined optical business, spanning interconnects, optical DSPs and Silicon Photonics, to exceed $500 million in revenue by fiscal 2027 1

1

. The deal is projected to close in the second quarter of this year, subject to customary closing conditions and regulatory approvals, and is expected to be accretive to non-GAAP EPS by fiscal 2027 1

1

2

2

. This acquisition builds on Credo's strategy of acquiring specialized tech companies to enhance AI data center capabilities, following its March acquisition of high-speed connectivity IP firm CoMira Solutions and its October 2025 Hyperlume acquisition 1

1

.

Industry Consolidation Mirrors Broader Trends

Credo's move reflects a broader pattern of consolidation in the semiconductor industry as companies race to capture AI infrastructure opportunities. Marvell Technology has similarly pursued acquisitions, including its 2018 Cavium deal that expanded its reach into networking and data centers 1

1

. In February 2026, Marvell Technology completed the acquisition of XConn Technologies to expand its PCIe and CXL switching portfolio, expecting $50 million annualized revenue in the fourth quarter of fiscal 2027 and $100 million in fiscal 2028 1

1

. DustPhotonics had raised over $100 million in venture capital funding from investors including Intel Capital, WRVI Capital, Greenfield Partners and Atreides Management before the acquisition 2

2

. As AI workloads continue to drive demand for faster, more efficient connectivity in data centers, companies with integrated solutions spanning both electrical and optical technologies are positioned to capture significant market share in the evolving landscape of hyperscale AI clusters.