POET Technologies Targets AI's Data Bottleneck with Advanced Photonics

TORONTO, ON – September 04, 2026 – As the artificial intelligence revolution accelerates, the digital highways carrying its data are straining under unprecedented traffic. Addressing this critical chokepoint, POET Technologies today announced its strategic participation in two of Asia's most influential technology conferences, where it will showcase its novel solutions for the AI and data center markets. The company will exhibit at the 2026 China International Optoelectronic Expo (CIOE) and present at the Infostone Optical Communication and Market Technology Conference (IFOC) in Shenzhen, China.

The move signals a concerted effort by the Toronto-based designer of Photonic Integrated Circuits (PICs) to capture a significant share of a market desperate for faster, more efficient, and scalable connectivity. At the heart of this push is the POET Optical Interposer™, a patented platform that enables the seamless, wafer-level integration of electronics and photonics. This technology promises to directly confront what industry experts call the "interconnect wall"—the point where the communication links between processors, not the processors themselves, become the primary bottleneck in advancing AI capabilities.

The AI Interconnect Challenge: A Multi-Billion Dollar Race

The insatiable data appetite of AI models is fueling an explosive expansion in the optical interconnect market. Projections show the market for AI data center optics rocketing from approximately $16.5 billion in 2025 to over $43 billion by 2035. This growth is driven by a frantic upgrade cycle, with hyperscale data center operators like Meta, Google, and Microsoft rapidly transitioning from 400G to 800G and now planning for 1.6T (terabits per second) and beyond.

This breakneck pace is a direct response to fundamental challenges. Traditional copper-based electrical connections are hitting their physical limits in terms of speed, distance, and especially power consumption. "AI is the main driver behind projections that see global data center electricity consumption nearly doubling between 2024 and 2030," noted one industry analyst. "Reducing the power budget of the interconnect fabric isn't just a matter of cost savings; it's becoming a prerequisite for building next-generation AI factories at scale."

This environment has created a fiercely competitive landscape. Established giants like Broadcom, Coherent, and Lumentum are investing billions to ramp up production of advanced optical components. The strategic importance of this sector was underscored in March when NVIDIA announced multi-billion-dollar investments and purchase commitments with both Coherent and Lumentum to secure its future supply of lasers and related technologies. The race is on to provide the high-density, low-power optical engines that will form the nervous system of future AI clusters.

POET's Answer: The Wafer-Level Optical Interposer

POET Technologies is positioning its Optical Interposer platform as a disruptive solution in this high-stakes race. The company's core innovation is a departure from traditional, complex methods of assembling optical components one by one. Instead, it utilizes advanced semiconductor wafer-level manufacturing techniques to integrate electronic and photonic devices into a single, compact package.

This approach, the company claims, yields optical engines that are lower in cost, smaller, and significantly more power-efficient. These are precisely the metrics that matter most to hyperscale data center operators struggling to manage capital and operational expenditures while scaling their AI infrastructure. The platform's alignment with Silicon Photonics (SiPh), a technology expected to command over 60% of the market's revenue by 2030, places it at the center of a major industry trend.

“For 2026, we expect to share details of our continued manufacturing progress and the reasons why a growing number of industry leaders see POET’s wafer-level chip-scale packaging technology as a viable solution to their needs for low-cost, high-power optical interconnects,” stated Dr. Suresh Venkatesan, POET Chairman and CEO. He added, “We will also be discussing the Company’s continued advancements in deploying next-generation applications for AI connectivity in hyperscale data centers.”

Dr. Venkatesan’s comments highlight a crucial advantage: scalability. By leveraging existing semiconductor fabrication infrastructure, the Optical Interposer platform is designed for high-volume production, a key factor in a market where demand consistently outstrips supply.

Strategic Showcase in Asia's Tech Hub

The decision to make a significant showing at CIOE and IFOC is a calculated strategic move. With over 3,800 exhibitors and an expected 240,000 professional visitors, CIOE is a premier global stage for the optoelectronics industry. Its location in Shenzhen places POET at the epicenter of the Asia-Pacific market, which is the world's fastest-growing region for data center infrastructure. China alone accounted for an estimated 43% of the regional market in 2025, making it an essential target for any company with global ambitions.

Preceding the main exhibition, the company’s technical prowess will be on display at IFOC. Dr. Mo Jinyu, Senior Vice President of Global Product Development, will deliver a presentation on “High Power and Multi-wavelength Laser Light Sources for CPO/AI/ML Interconnects.” This topic is far from academic; it addresses the fundamental physics of how to transmit more data, more efficiently. High-power lasers are needed to ensure signal integrity across larger AI clusters, while multi-wavelength technology—using different colors of light to carry separate data streams down a single fiber—is the key to unlocking the massive bandwidth required for 1.6T and 3.2T speeds.

This focus on advanced light sources for Co-Packaged Optics (CPO)—the integration of optics directly onto the same package as the main AI processor—demonstrates that the company is aiming its technology squarely at the industry's next frontier for solving latency and power consumption problems.

From Prototype to Profit: The Path to Commercialization

For any technology company, the journey from an innovative prototype to profitable, high-volume production is the ultimate test. POET’s presence in Shenzhen is about more than just demonstrating technical capabilities; it's a critical milestone on its commercialization path. The “continued manufacturing progress” mentioned by its CEO is the lynchpin of this journey. Recent developments, including a reported $50 million purchase order for its Lumilens light source products from a hyperscaler supplier and partnerships with manufacturing and development firms like LITEON and Lessengers, provide tangible evidence that the company is transitioning from development to revenue generation.

By engaging directly with the vast ecosystem of customers, partners, and suppliers at CIOE, the company is actively building the relationships needed to scale its operations. This is happening at an opportune moment. Hyperscale operators, wary of supply chain concentration and hungry for innovation that can provide a competitive edge, are actively evaluating new suppliers. For companies like POET that can prove their technology is not only advanced but also manufacturable at scale and cost-effectively, the window of opportunity to power the future of artificial intelligence is wide open.