CScale Raises US$145 Million Series C for Resilient AI Optical Interconnects
  • News
  • North America

CScale Raises $145 Million Series C for Resilient AI Optical Interconnects

Round brings total funding to US$188 million as NVIDIA and Intel Capital join

10/1/2026
•Ghita Khalfaoui
Back to News

CScale, a developer of optical interconnect technology for artificial intelligence infrastructure, emerged from stealth on September 30, 2026 with a US$145 million Series C round. The funding brings the company's total capital raised to US$188 million as it works to commercialize high bandwidth, low latency connectivity for AI scale-up systems. The investment signals growing demand for more reliable communication layers inside next-generation AI data centers.


Series C Funding and Strategic Backers

The Series C round was led by Atreides Management, with Valor Equity Partners and Premji Invest joining as co-leads. Sutter Hill Ventures, which has supported CScale since its inception, continued its participation alongside existing investor Maverick Silicon. NVIDIA and Intel Capital joined the round as CScale's first strategic investors, reflecting interest from established chip and data center technology leaders.

Reliability at Gigawatt Scale

CScale argues that gigawatt-class AI data centers expected by the end of the decade will require scale-up domains spanning thousands of tightly coupled accelerators across dozens of racks. In such environments, rare failures on individual optical links become a continuous fleet-level condition that can degrade overall AI factory performance. The company is designing its interconnect to contain optical failures without interrupting compute, prioritizing continuity over simple part replacement.

Executive Leadership and Company Mission

CScale is led by Chief Executive Officer Martin Lund, who previously built Broadcom's switching business into a billion-dollar franchise and held senior roles at Microsoft, Cadence, and Cisco. Founder and Chief Technology Officer Sanjai Kohli has created and exited multiple communications startups, including SiRF, which helped bring GPS to the mass market. The company was founded in 2023 and now employs approximately 85 people globally from its headquarters in Palo Alto, California.

Investor Confidence in Systems-Level Design

Atreides Management Managing Partner and Chief Investment Officer Gavin Baker described AI infrastructure as a systems problem rather than a purely compute problem. He said interconnect bandwidth has little value without system reliability, and that optical failures represent compute failures at very large scale. Baker added that CScale is designing reliable scale-up optics for next-generation AI factories without tradeoffs.

Premji Invest Perspective

Premji Invest Managing Partner Sandesh Patnam praised CScale's architectural judgment and its ability to identify technical choices that create value across the entire system. He noted the team combines expertise in photonics, electronics, and software with deployment experience needed to turn optical capability into dependable AI infrastructure. Patnam said Premji Invest is proud to co-lead the financing and support CScale's transformative solutions.

About CScale

CScale is building an integrated light engine to deliver reliable optical connectivity for AI infrastructure at gigawatt scale. Lund previously led Cisco's Common Hardware Group, where he was responsible for silicon, hardware systems, and optics including Silicon One. Kohli received the 2010 European Inventor Award and later founded Inovi, which Facebook acquired in 2014.


The new capital positions CScale to accelerate development and commercialization of optical interconnect designed for the extreme reliability requirements of modern AI factories. By combining strategic backing from NVIDIA and Intel Capital with long-term investors such as Sutter Hill Ventures, the company is aligning itself with key players in AI infrastructure. Its focus on continuous operation during optical failures addresses a practical challenge that will become more important as AI clusters grow.