The AI infrastructure race is heating up, and two photonics players are taking aim at the next bottleneck—speed. POET Technologies (NASDAQ: POET) and Quantum Computing Inc. (NASDAQ: QUBT) have announced a strategic collaboration to develop 400G/Lane thin-film lithium niobate (TFLN) modulator-based 3.2Tbps optical engines, setting their sights on the future of ultra-high-performance AI networking.
The partnership brings together POET’s Optical Interposer™ platform, which integrates photonic and electronic components on a single chip, and QCi’s expertise in TFLN-based modulators—a material and method increasingly seen as a breakthrough in photonic speed and efficiency. Together, they aim to deliver next-generation optical engines that power the world’s most data-hungry AI systems and hyperscale data centers.
A New Wave in Optical Connectivity
In plain terms, POET and QCi want to double the current state of the art. Today’s fastest optical engines top out around 1.6Tbps. By integrating 400G/Lane modulators into POET’s platform, the new 3.2Tbps design promises a twofold jump in throughput, providing the kind of backbone needed for the ballooning compute demands of large AI models and high-density cloud workloads.
“To achieve a commercially viable optical modulator that operates at 400G/Lane requires an integration platform such as the POET Optical Interposer,” said Dr. Yong Meng Sua, CTO of QCi. “Combining our TFLN designs with POET’s platform addresses the global demand for more compute power.”
The companies expect the development to complete in the second half of 2026, with POET funding the project.
The Photonics Power Couple
TFLN—short for thin-film lithium niobate—has become one of the most promising materials in the optical industry. Known for its resilience, ultra-fast response, and low power loss, it’s ideal for next-gen modulators that convert electronic signals into light at blistering speeds. QCi is among a handful of companies capable of adapting TFLN for wafer-level manufacturing, a key step toward commercial viability.
POET’s Optical Interposer™, meanwhile, acts as the glue connecting optical and electrical domains on a single platform—reducing complexity, cutting costs, and improving scalability. Together, these technologies could form the foundation of next-generation photonic compute fabrics—critical for AI workloads, edge computing, and cloud-scale data centers.
The Stakes: AI’s Thirst for Bandwidth
The timing couldn’t be better. As generative AI and high-performance computing architectures proliferate, data movement—not processing power—is increasingly the limiting factor. Even the most advanced GPUs and NPUs rely on high-speed optical links to communicate efficiently within massive clusters.
“The work we are doing with QCi keeps POET at the forefront of AI connectivity development,” said Dr. Suresh Venkatesan, POET’s Executive Chairman and CEO. “A 400G/Lane optical modulator that incorporates TFLN will be a radical step forward for the industry.”
According to LightCounting’s 2025 ‘Ethernet Optics: Cloud, Enterprise and Telecom’ report, the global market for 3.2T pluggable and co-packaged optics (CPO) could reach nearly $12 billion by 2030—a lucrative horizon for any player that can deliver speed and scalability at once.
With hyperscalers like Google, Amazon, and Microsoft investing heavily in AI-driven infrastructure, demand for energy-efficient, high-bandwidth optical solutions is skyrocketing. POET and QCi’s partnership reflects a broader industry pivot toward photonic integration, where light replaces copper for data transport at every level—from chips to racks to entire data centers.
If the collaboration succeeds, it could not only accelerate AI networking speeds but also set the tone for co-packaged optics and photonic compute interconnects that underpin the next decade of computing innovation.
In a world where even GPUs are waiting for bandwidth, POET and QCi are betting on light to lead the way.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
