Scintil Photonics, a French deep-tech startup originating from CEA-Leti, has closed a $58 million Series B funding round, underscoring growing investor confidence in integrated photonics as a cornerstone of next-generation AI infrastructure. The round was co-led by Yotta Capital Partners and NGP Capital, with NVIDIA joining as a new strategic participant. Additional backing came from BNP Paribas Développement, while existing investors—including Supernova Invest, Bpifrance Digital Venture, Innovacom, Bosch Ventures, and Applied Ventures ITIC Innovation Fund (AVITIC)—also contributed.
The investment marks a major milestone for Scintil, which is positioning itself at the nexus of photonics, AI, and data center scalability. With AI workloads demanding unprecedented bandwidth and energy efficiency, Scintil’s breakthrough technology offers a pathway to higher performance, lower costs, and reduced environmental impact.
SHIP technology: a breakthrough in integration
At the core of Scintil’s innovation is its proprietary SHIP (Scintil Heterogeneous Integration Photonics) process, which integrates lasers, modulators, and photodiodes on a single chip. Traditionally, these optical components exist as separate devices, adding cost, complexity, and power overhead. By consolidating them, Scintil achieves:
Higher density: Dozens of discrete parts condensed into a monolithic chip solution.
Greater efficiency: Optimized power usage, reducing datacentre energy draw.
Improved scalability: A design that can be manufactured at high volume using a commercial supply chain.
This integrated approach promises performance levels previously unattainable with conventional photonics, delivering both technical and economic advantages for hyperscale AI infrastructure.
LEAF Light: powering AI factories
Scintil’s Series B funding will accelerate the commercial ramp of LEAF Light, the company’s flagship product. Positioned as the industry’s first DWDM-native light engine tailored for co-packaged optics (CPO), LEAF Light offers a disruptive combination of bandwidth density and energy savings.
Key performance highlights include:
6.4T/mm edge bandwidth density, enabling unprecedented optical throughput.
DWDM-native design, outputting multiple precisely spaced wavelengths from a single chip, vastly increasing bandwidth.
Energy efficiency: Operating at roughly one-sixth the power consumption of pluggable optics, lowering operational costs while reducing carbon footprint.
Designed for scale-up GPU clusters and AI datacenters, LEAF Light enables low-power, high-density optical connectivity, a critical enabler for companies building the world’s most advanced AI factories.
CEO Matt Crowley emphasized the strategic leap:
“Our SHIP technology enables integrated photonic solutions with the scalability, energy efficiency, and integration density required to power next-generation compute infrastructure. This efficiency not only reduces datacentre operating costs but also contributes to lowering the carbon footprint of AI infrastructure. With LEAF Light entering high-volume production, we’re expanding from our base in Grenoble into the international markets, including the US, to support the world’s most advanced AI factories.”
Customer-driven innovation
Scintil has developed LEAF Light in close collaboration with hyperscale customers, ensuring alignment with the real-world requirements of AI compute providers. Founder and CTO Sylvie Menezo described the breakthrough as a monolithic integration of distributed feedback (DFB) lasers onto silicon photonic circuits, a feat achieved while retaining manufacturability through established supply chains.
She noted:
“Used as an external laser source for co-packaged optics (CPO) transmissions, LEAF Light is a key component for the next generation of AI datacentres. Its uniqueness lies in a single-chip solution that monolithically integrates precisely spaced DFB lasers on silicon photonic circuits. Our SHIP technology provides a strong and disruptive foundation to expand our portfolio of integrated photonic circuit solutions for the coming decades.”
Strategic investor validation
The participation of NVIDIA in the round is especially notable, as the chip giant continues to lead in GPU acceleration for AI while seeking complementary technologies to manage the interconnect bottleneck. As GPU clusters scale, optical interconnects are becoming mission-critical, and NVIDIA’s involvement signals strong confidence in Scintil’s potential to deliver.
Lead investor Vincent Deltrieu of Yotta Capital Partners highlighted Scintil’s alignment with pressing AI infrastructure needs:
“Scintil exemplifies the kind of innovation leaders we look for, combining advanced manufacturing, deep-tech leadership, and meaningful impact on the energy demands of AI infrastructure. Scintil’s integrated photonics platform is essential to scale the next generation of AI factories.”
Global expansion
With Series B funding in place, Scintil plans to:
Expand hiring in France, the U.S., and internationally.
Scale production capacity to meet growing customer demand.
Deepen international presence with new commercial operations.
By transitioning from R&D leadership to full-scale deployment, Scintil is positioning itself as a critical enabler of global AI infrastructure, combining photonics integration with datacentre-ready performance.
Scintil Photonics enters the high-stakes race for AI datacenter optics
The $58 million Series B funding marks a critical juncture for Scintil Photonics, positioning the Grenoble-based startup in one of the most competitive technology markets: integrated photonics for AI datacenter interconnects. With demand for bandwidth and energy efficiency reaching unprecedented levels, Scintil now finds itself competing with global players such as Ayar Labs, Ranovus, Intel, and Cisco, each racing to deliver scalable optical solutions for hyperscale GPU clusters.
Why optics are critical for AI
The rise of generative AI and large-scale machine learning has transformed datacenter architecture. Traditional electrical interconnects are increasingly unable to handle the:
Bandwidth density requirements of GPU-to-GPU communication.
Latency sensitivity of distributed AI training workloads.
Energy efficiency constraints in facilities already consuming gigawatts of power.
As GPUs become more powerful, the interconnect bottleneck has emerged as a limiting factor. Optical connectivity, particularly co-packaged optics (CPO) and dense wavelength division multiplexing (DWDM), is seen as the only viable pathway to sustaining AI performance growth.
Scintil’s differentiator: SHIP + DWDM-native integration
While many players focus on optical engines and pluggable modules, Scintil brings a distinctive two-pronged advantage:
SHIP process technology: Monolithic integration of lasers, modulators, and photodiodes onto a single chip. This collapses traditionally bulky optical assemblies into compact, manufacturable circuits.
DWDM-native design: LEAF Light outputs multiple precisely spaced wavelengths directly from the chip, dramatically expanding bandwidth without proportional increases in power or physical footprint.
The result is 6.4T/mm edge bandwidth density, delivered at roughly one-sixth the power consumption of conventional pluggable optics—performance metrics few competitors can claim today.
Key competitors
Ayar Labs (U.S.) – Backed by Intel and NVIDIA, Ayar has pioneered optical I/O using micro-ring modulators. Its TeraPHY chiplets are already being tested in HPC and AI systems, giving it a head start in customer adoption.
Ranovus (Canada) – Known for its ODIN optical engines, which integrate multi-wavelength lasers and modulators, Ranovus is working with IBM and other partners to scale datacenter optics.
Intel (U.S.) – A long-time champion of silicon photonics, Intel offers pluggable and integrated optical modules, leveraging its semiconductor scale for high-volume manufacturing.
Cisco (U.S.) – Through its acquisitions in optics (Acacia, Luxtera), Cisco is embedding photonics into its networking gear, giving it downstream influence in datacenter architectures.
Scintil competes by emphasizing monolithic integration and DWDM-native capability, where many rivals still rely on hybrid assembly or separate laser sources. This could provide Scintil with an efficiency edge and cost advantage as customers scale toward exascale GPU clusters.
European angle: sovereignty in photonics
Scintil also plays into Europe’s strategic ambition to reduce reliance on U.S. and Asian technology providers in critical compute infrastructure. By building on CEA-Leti research and maintaining a European supply chain, Scintil contributes to regional resilience in semiconductor and photonics manufacturing.
The participation of NVIDIA in its Series B highlights an important dynamic: U.S. hyperscalers may prefer partnering with local innovators to secure supply diversity and mitigate geopolitical risks. Scintil could thus become a bridge between European R&D excellence and global AI infrastructure demand.
Market outlook: explosive growth
The market for datacenter optics is forecast to grow from $12 billion in 2023 to more than $30 billion by 2030, driven primarily by AI training and inference clusters. Within this, co-packaged optics is expected to shift from pilot adoption in 2025 to mainstream deployment by 2027.
Scintil’s LEAF Light, already aligned with CPO architectures, positions the company well to capture early deployments at hyperscalers and GPU manufacturers. Its combination of low power, high density, and manufacturability directly addresses the top concerns of datacenter operators.
Challenges ahead
Despite its strong positioning, Scintil faces hurdles:
Commercial scaling: Transitioning from prototypes to high-volume manufacturing requires flawless supply chain execution.
Competitive timing: U.S. rivals like Ayar Labs may lock in partnerships with NVIDIA, AMD, and Intel earlier, creating barriers for Scintil’s entry.
Customer validation: Hyperscalers will demand extensive reliability testing before adopting new photonic technologies at scale.
Capital intensity: Integrated photonics is highly capital-intensive, and Scintil may require further rounds of funding to sustain global expansion.
Strategic implications
By securing Series B funding and attracting NVIDIA as a strategic investor, Scintil has signaled its readiness to compete on the global stage. Its ability to deliver DWDM-native single-chip light engines could reshape how bandwidth and energy efficiency are engineered into AI datacenters.
If Scintil executes its roadmap, it may not only challenge North American rivals but also position Europe as a leader in integrated photonics for AI infrastructure—a sector that will define the performance and sustainability of digital economies over the next decade.
Powered by Froala Editor