Xsight Labs Raises $300M for Open Chips That Link GPUs in AI Clusters

Israeli company Xsight Labs has raised more than $300M at a $2.8B valuation in a round led by Fidelity. That is more than a fivefold jump in valuation. Xsight makes open networking chips that connect GPUs in AI clusters, a segment that is quietly becoming as competitive as the accelerator market itself.

AI-processed from TNW; edited by Hamidun News
Xsight Labs Raises $300M for Open Chips That Link GPUs in AI Clusters
Source: TNW. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceTNW ↗
Editor-in-Chief
Key metric #1$300M+ — total funding raised
Key metric #2$2.8B — the company's current valuation
Key metric #3More than 5x — increase in valuation over the previous round

Israeli chipmaker Xsight Labs announced on July 31, 2026 that it has raised more than $300M at a $2.8B valuation, more than five times its previous valuation. Investment firm Fidelity led the round.

What Xsight Labs Makes

Xsight Labs develops open networking chips that connect thousands of GPUs into single compute clusters for AI. While the world's attention has been fixed on accelerator makers, the unglamorous networking layer — high-speed switches and interconnects — has become equally critical and expensive infrastructure in AI data centers.

Key facts about the round:

  • $300M+ — total funding raised
  • $2.8B — the company's current valuation
  • More than 5x — increase in valuation over the previous round
  • Fidelity — lead investor
  • Israel — where the company was founded and develops its products

Why Have Networking Chips Become a Strategic Asset?

Without high-speed networking chips, thousands of GPUs cannot operate as a single supercomputer: moving data between accelerators creates a bottleneck that cancels out the compute potential of expensive hardware. That is the problem Xsight Labs solves.

"The boring chips that connect thousands of GPUs to each other are quietly becoming as valuable as the accelerators themselves,"

The Next Web writes, describing the shift in the AI infrastructure market.

The word "open" in the company's positioning is strategically key. NVIDIA, with its proprietary InfiniBand technology, largely controls the market for AI networking chips today. Xsight's open architecture lets cloud providers and AI labs build clusters without being locked into a single vendor. For hyperscalers — Amazon, Google, Microsoft and the new AI companies — it offers a real chance to diversify infrastructure purchasing and reduce their dependence on NVIDIA.

As data centers scale GPU density from thousands to tens of thousands of accelerators, networking costs in a project's budget are becoming comparable to the cost of the GPUs themselves, which explains the sharp rise in investor interest in this segment.

With GPU export restrictions and periodic accelerator shortages now permanent features of the market, suppliers of open networking solutions gain an additional strategic advantage.

Israel as a Hub for Chip Engineering

Xsight Labs is part of a growing wave of Israeli semiconductor startups in AI infrastructure. Israel has long been one of the world's centers of chip engineering: Intel, NVIDIA and Qualcomm all maintain large R&D divisions there, which creates a dense pool of engineering talent for a semiconductor startup.

The $2.8B valuation makes Xsight Labs one of the most highly valued independent players in AI networking infrastructure. The fivefold jump from the previous round signals that institutional investors are betting on a sharp scale-up in demand for precisely this class of components.

What This Means

The Xsight Labs round confirms that the AI infrastructure race has moved well beyond GPUs. Networking chips, long overshadowed by accelerators, are turning into a strategic asset in their own right, with multibillion-dollar valuations. Xsight's open architecture is a direct challenge to NVIDIA's dominance in interconnects, and major providers are actively looking for alternatives to proprietary solutions.

Frequently Asked Questions

What are open networking chips?

Open networking chips are components for connecting GPUs in a cluster that are built on open standards. Unlike NVIDIA's proprietary InfiniBand technology, they let companies build high-speed networks without being tied to one manufacturer's hardware.

How much is Xsight Labs worth after the round?

According to The Next Web, the company was valued at $2.8B following the round completed on July 31, 2026, more than five times its previous valuation. Fidelity was the lead investor.

Story

Key Questions Answered

Why Have Networking Chips Become a Strategic Asset?

Without high-speed networking chips, thousands of GPUs cannot operate as a single supercomputer: moving data between accelerators creates a bottleneck that cancels out the compute potential of expensive hardware. That is the problem Xsight Labs solves. "The boring chips that connect thousands of GPUs to each other are quietly becoming as valuable as the accelerators themselves," The Next Web writes, describing the shift in the AI infrastructure market. The word "open" in the company's positioning is strategically key. NVIDIA, with its proprietary InfiniBand technology, larg

What are open networking chips?

Open networking chips are components for connecting GPUs in a cluster that are built on open standards. Unlike NVIDIA's proprietary InfiniBand technology, they let companies build high-speed networks without being tied to one manufacturer's hardware.

How much is Xsight Labs worth after the round?

According to The Next Web, the company was valued at $2.8B following the round completed on July 31, 2026, more than five times its previous valuation. Fidelity was the lead investor.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Unified AI Workspace · Nomad

Test Frontier LLMs & Autonomous Agents in One Workspace

Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.

Try Nomad AI →Launch Nomad Agent
What do you think?
Loading comments…