Marvell to acquire Celestial Artificial Intelligence for photonic fabric

Marvell Technology, Inc. has entered into a definitive agreement to acquire Celestial Artificial Intelligence, a developer of a Photonic Fabric platform aimed at scale-up optical interconnect for next-generation data centers.

Marvell Technology, Inc. (NASDAQ: MRVL) announced it has entered into a definitive agreement to acquire Celestial Artificial Intelligence, a company described as a pioneer of a Photonic Fabric technology platform purpose-built for scale-up optical interconnect. Marvell framed the move as a strategic acquisition that represents a transformational milestone in accelerating its connectivity strategy for next-generation Artificial Intelligence and cloud data centers.

Celestial Artificial Intelligence’s Photonic Fabric platform is presented as a disruptive technology intended to enable high-bandwidth, ultra-low latency optical interconnects at scale. The article highlights that the platform is purpose-built for scale-up optical interconnect, positioning it to support multi-rack architectures and to serve as an integrated fabric between processing units. Marvell’s announcement frames the technology as complementary to its data infrastructure semiconductor solutions portfolio.

The article situates the acquisition within a broader shift in data center architecture driven by Artificial Intelligence. It explains that next-generation accelerated systems are evolving beyond single-rack configurations into multi-rack systems that connect hundreds of XPUs with an any-to-any scale-up fabric. That architecture is described as allowing each XPU to access the memory of every other XPU directly. The piece notes that these advanced fabrics demand purpose-built switches and protocols such as UALink, engineered to deliver the performance and efficiency required at scale. Marvell casts the deal as advancing its connectivity strategy to address those emerging architectural needs in Artificial Intelligence and cloud environments.

58

Impact Score

Samsung completes hbm4 development, awaits NVIDIA approval

Samsung says it has cleared Production Readiness Approval for its first sixth-generation hbm (hbm4) and has shipped samples to NVIDIA for evaluation. Initial samples have exceeded NVIDIA’s next-gen GPU requirement of 11 Gbps per pin and hbm4 promises roughly 60% higher bandwidth than hbm3e.

NVIDIA and AWS expand full-stack partnership for Artificial Intelligence compute platform

NVIDIA and AWS expanded integration around Artificial Intelligence infrastructure at AWS re:Invent, announcing support for NVIDIA NVLink Fusion with Trainium4, Graviton and the Nitro System. the move aims to unify NVIDIA scale-up interconnect and MGX rack architecture with AWS custom silicon to speed cloud-scale Artificial Intelligence deployments.

the state of artificial intelligence and DeepSeek strikes again

the download highlights a new MIT Technology Review and Financial Times feature on the uneven economic effects of Artificial Intelligence and a roundup of major technology items, including DeepSeek’s latest model claims and an Amsterdam welfare Artificial Intelligence investigation.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.