Back to News
IndustryAI Understanding briefing

CScale raises $145 million Series C to build optical interconnects for AI data centers

Palo Alto‑based CScale emerged from stealth with a $145 million Series C round led by Atreides Management and Valor Equity Partners, aiming to commercialise optical links that keep AI accelerators running despite laser failures.

4 min readRead the linked source
Source-page capture accompanying CScale raises $145 million Series C to build optical interconnects for AI data centers
Source referenceSource recorded
Publisher
techfundingnews.com
Source link
techfundingnews.comhttps://techfundingnews.com/optical-ai-chip-startup-cscale-emerges-from-stealth-with-145m-funding/
Source type
Linked source — primary-source status has not been established.
ContextUnderstand this in 60 seconds

Start here

Key terms

Inference
The runtime phase where a trained model generates predictions or outputs.
Compute
The processing resources required to train and run models, often measured in FLOPS or GPU hours.
Latency
The time between sending a request and receiving the model's output.
Test yourselfFuture of AI Quiz

What happened

CScale announced a $145 million Series C financing round on September 30, according to techfundingnews.com. The round was co‑led by Atreides Management and Valor Equity Partners, with Premji Invest joining as a co‑lead investor. Existing backers Sutter Hill Ventures and Maverick Silicon continued their support, while NVIDIA and Intel Capital participated as the startup’s first strategic investors. The new capital brings total funding to $188 million and is earmarked for development and commercialisation of CScale’s optical interconnect platform designed for large‑scale AI data‑centre deployments.

Techfundingnews.com reported that CScale, founded in 2023, emerged from stealth on September 30 with a $145 million Series C round. The round was led by Atreides Management and Valor Equity Partners, with Premji Invest co‑leading. Existing investors Sutter Hill Ventures and Maverick Silicon continued their participation, while NVIDIA and Intel Capital joined as the first strategic investors. The financing brings CScale’s total capital raised to $188 million.

The funding is earmarked for the development and commercialisation of CScale’s optical interconnect platform, which the company describes as an integrated light engine capable of reliable, ultra‑low‑ communication between AI accelerators. The platform is designed to contain laser failures without interrupting computation, a problem that becomes acute in gigawatt‑scale AI data centres where thousands of links operate simultaneously.

CScale’s leadership includes CEO Martin Lund, who brings three decades of experience from Broadcom, Microsoft, Cadence, and Cisco, and CTO Sanjai Kohli, a co‑founder of GPS‑chip maker SiRF and former founder of Inovi (acquired by Facebook). Their combined expertise underpins the company’s focus on both silicon and optical networking.

Source details: techfundingnews.com ↗

Why it matters

The AI industry is rapidly moving toward "scale‑up" architectures that stitch together thousands of accelerators across dozens of racks to achieve gigawatt‑class power. Traditional copper interconnects struggle with bandwidth and at that scale, while optical links promise higher throughput and lower latency. However, optical components such as lasers can fail, and a single failure can cascade and halt computation. CScale’s approach—building an integrated light engine that contains failures and maintains continuity—addresses a critical reliability gap for next‑generation AI factories. The involvement of strategic investors like NVIDIA and Intel Capital signals strong industry confidence that optical interconnects will become a core infrastructure layer for AI workloads, potentially reshaping data‑centre design and economics.

Optical interconnects are increasingly viewed as essential for the next generation of AI infrastructure because they can deliver the bandwidth and required for massive model training and . CScale’s emphasis on reliability addresses a known weakness of photonic links—laser failure—that could otherwise negate performance gains.

The participation of NVIDIA and Intel Capital as strategic investors suggests that leading AI hardware manufacturers see value in integrating CScale’s technology into their future product roadmaps. This could accelerate the adoption of optical interconnects across the broader AI ecosystem, influencing data‑centre design, power consumption, and total cost of ownership.

CScale’s funding round also reflects a broader trend of substantial capital flowing into photonic‑AI startups. Recent large rounds for Celestial AI, Lightmatter, and Eliyan indicate a competitive market where multiple approaches to optical connectivity are being pursued, potentially leading to rapid innovation and standardisation.

Interactive Mechanism

Interactive Mechanism: How It Actually Works

Explore the underlying technology behind this development interactively.

Model Parameter Size:8B Parameters
VRAM Required5.5 GBGPU memory footprint
Target HardwareMacBook / Single GPUDeployment tier
Privacy100% Air-GappedLocal device capability
Core takeaway: Small, quantized models (3B–8B) now run directly inside smartphones and laptops with complete data privacy, while mammoth 400B+ models remain the domain of datacenter clusters.
Interactive Concept Check+10 Points
Future of AI Quiz

What did neural scaling law research (e.g. Kaplan et al., 2020) observe?

What to watch next

Key indicators to monitor include: (1) CScale’s prototype demonstrations and any announced partnerships with AI hardware vendors; (2) timelines for product availability and pricing, which have not been disclosed; (3) adoption signals from large‑scale AI cloud providers that may pilot the technology; and (4) competitive moves from other photonic‑interconnect startups such as Celestial AI, Lightmatter, and TFN’s Eliyan, which could accelerate market consolidation.

Prototype demonstrations: CScale’s ability to showcase a working optical link that tolerates laser failures will be a key proof point for investors and potential customers.

Partnership announcements: Any formal collaborations with major AI hardware vendors (e.g., NVIDIA, Intel, AMD) would signal a path to large‑scale deployment.

Product timeline and pricing: The company has not disclosed when its interconnect solution will be commercially available or at what cost, leaving uncertainty about market accessibility.

Competitive landscape: Progress from rivals such as Celestial AI, Lightmatter, and Eliyan could influence CScale’s market positioning and the pace of industry adoption.

Related guides & quizzes

Future of AIAI TrainingAI Models ExplainedTest what you know — try a free AI quizLook up an AI term in our glossaryFollow the AI funding tracker
Found this useful?