Optical Interconnect Startup CScale Secures $145 Million
Data center operators may soon gain access to new hardware designed to prevent system-wide compute failures.
Updated on Oct. 26, 2026 in Startups
Live Poll
Do you believe current AI infrastructure investments are successfully prioritizing system reliability over raw performance?
Startup CScale has exited stealth mode with $145 million in new Series C funding. The company develops optical interconnects designed to isolate hardware failures within AI compute clusters.
Why it matters
As AI processing demands grow, optical interconnect reliability has become a primary bottleneck for data centers. This funding allows CScale to accelerate the commercialization of its hardware, which aims to keep AI compute systems running during localized optical failures.
The firm secured $145 million in its latest funding round, bringing its total capital raised to $188 million since its 2023 founding. Future production capacity and specific product launch dates remain unknown.
The players
CScale
A Palo Alto-based startup developing optical interconnect technology for AI compute clusters.
Sanjai Kohli
The founder of CScale, who launched the company in 2023.
Nvidia
A primary developer of AI processors that joined the round as a strategic investor.
Intel Capital
The venture capital arm of the semiconductor manufacturer that participated as a strategic investor.
The details
CScale specializes in optical interconnects designed to manage high-speed data transmission between AI processors. The technology operates by isolating optical failures, ensuring that a single component malfunction does not force a shutdown of the entire compute cluster. The company will use the capital to scale manufacturing and accelerate its commercial roadmap.
Timeline
2023: CScale is founded by Sanjai Kohli.
October 2026: The company exits stealth mode.
Market Landscape
Investment in optical interconnects for AI has surged as data center operators seek to solve latency and reliability bottlenecks in large-scale compute. This financing follows a broader industry trend of semiconductor leaders betting on infrastructure-level hardware to support escalating AI workloads.
Operators managing AI compute infrastructure should monitor CScale's commercial release schedule as a potential signal for improving cluster uptime. Procurement teams should evaluate how optical redundancy hardware could reduce downtime costs compared to current standard interconnect solutions.
The takeaway
CScale's entry highlights how deep-tech startups are moving from research to infrastructure supply for AI data centers. Operators should track how the integration of redundant optical interconnects impacts their total cost of ownership and cluster availability metrics.
Further reading
For broader trends in infrastructure funding, read more in Startups.
Source note: This article includes information reported by Electronics Weekly.
Live Poll
Do you believe current AI infrastructure investments are successfully prioritizing system reliability over raw performance?
