Programmable networking chip startup Xsight Labs raises $300M

What happened

Israeli startup Xsight Labs Ltd. secured $300 million in a funding round that values the company at $2.8 billion. The capital injection will fuel expansion of its programmable Ethernet switches and 800-gigabit data processing units (DPUs), targeting cloud infrastructure and AI network applications. Fidelity led the investment round.

Why it matters

Xsight Labs is betting on programmable networking chips to reshape traffic management and processing at the data center edge. Its 800-gigabit DPUs suggest a push toward handling increasingly high-bandwidth AI workloads directly on the network, reducing latency and offloading tasks from CPUs and GPUs. The large funding round and near $3 billion valuation emphasize investor confidence in programmable networking hardware as a lever to improve cloud efficiency and control.

This move tightens competition in the networking chip market, where incumbents like NVIDIA and Intel also push DPUs and programmable switches. Buyers and cloud operators should expect intensifying innovation and pricing pressure as startups like Xsight Labs try to carve out slices in AI cloud infrastructure. Programmability in Ethernet switches opens paths for more customized, software-driven network management, potentially accelerating deployment of AI at scale.

What to watch next

Monitor how Xsight Labs positions itself against established chip vendors and whether it secures high-profile cloud or AI service provider customers. The startup’s ability to deliver robust programmable switch silicon and DPUs will determine if it can shift power toward more specialized networking hardware in AI workloads.

Investors will also watch for further funding rounds or partnerships that extend Xsight Labs’ market reach. On the technical side, look for real-world benchmarks showing reduced cloud latency or improved AI throughput using its chips. These outcomes will influence adoption by builders and operators seeking to optimize AI infrastructure costs and performance.

AI Quick Briefs Editorial Desk