Modern AI systems rely on thousands of processors, memory devices and networking components operating as a single computational platform. Training a large model requires continuously distributing data, exchanging intermediate results and synchronizing calculations across these devices. In traditional systems, these connections are made with copper traces, cables and electrical transceivers. At relatively short distances, electrical links remain efficient and inexpensive. However, as data rates rise, their physical limitations become increasingly severe. The resistance of conductors converts electrical energy into heat, while capacitance slows the charging and discharging of signals. At higher frequencies, skin effect, dielectric loss and signal reflections further distort the data stream.

The result is a growing communication bottleneck between processors. To compensate for signal degradation, electrical systems often require equalizers, retimers, amplifiers and other signal-conditioning circuits. These components consume power and add latency, particularly when a signal must pass through multiple stages. Electrical links also become more difficult to scale when many high-speed channels are placed close together, because neighbouring wires can interfere with one another through electromagnetic coupling. The problem is especially acute in AI hardware, where the performance of an accelerator is increasingly determined not only by its arithmetic capability but also by how quickly it can access memory and exchange data with other accelerators.

Optical interconnects approach the problem differently by transporting information as light. An electrical data signal is converted into an optical signal by a modulator, transmitted through a waveguide or optical fibre, and then converted back into an electrical signal by a photodetector. Light does not experience the same resistive losses as an electrical current flowing through a metal conductor, and optical channels can carry extremely high data rates over compact physical paths. Multiple wavelength channels can also share the same waveguide through wavelength-division multiplexing, allowing several independent data streams to travel simultaneously. These properties make optical communication attractive not only for long-distance data-centre networks but also for connections between chips in the same computing package.

Co-packaged optics brings the optical transceiver directly beside the electronic switching or computing device rather than placing it at the end of a long electrical connection. In a conventional architecture, a high-speed electrical signal may travel from a processor or switch across a package, through a board and into a separate optical module. Every additional millimetre of electrical routing increases loss and places greater demands on the transmitter and receiver. By positioning photonic components next to the main chip, co-packaged optics can shorten the electrical path and move the conversion from electrical to optical signals closer to the point where data is generated. The optical signal can then travel across the system with lower propagation loss and potentially lower energy per transmitted bit.

The review identifies three interconnected technological domains that determine whether optical compute interconnects can deliver system-level benefits. The first is the electrical subsystem, which includes drivers, receivers, serializers and deserializers, clocking circuits and power-management components. The second is the electro-optical and opto-electronic interface, where modulators encode data onto light and photodetectors recover it. The third is the optical transmission network, consisting of waveguides, fibres, couplers, switches and multiplexing structures. Improvements in only one domain may not be enough. A highly efficient optical modulator, for example, cannot compensate for inefficient driver electronics, poor coupling or excessive thermal overhead elsewhere in the package.

The roadmap described in the study follows the evolution of packaging from two-dimensional arrangements to increasingly integrated architectures. In a 2D co-packaged design, optical engines and electronic chips are placed side by side on a common substrate or package. This approach is comparatively accessible because it builds on established assembly methods, but it can require long connections across the package and may limit the density of optical channels. A 2.5D architecture uses an interposer, a specialised layer that provides dense electrical and optical routing between chiplets. By bringing compute, memory, switching and photonic components closer together, interposers can support shorter links and higher bandwidth densities while preserving some manufacturing flexibility.

The most ambitious stage is three-dimensional heterogeneous integration, in which different materials and devices are stacked vertically. Electronic logic, photonic circuits, laser sources and optical detectors could occupy separate layers optimized for their individual functions. Such integration could dramatically reduce the distance that signals travel and increase the number of connections available within a small footprint. It also introduces difficult engineering problems. Photonic devices often require materials and fabrication processes that differ from those used for advanced logic circuits. Aligning optical structures across stacked layers demands extreme precision, while repairing or testing a multilayer assembly can be more complicated than testing separate components.

Heat is another central challenge. Optical communication may reduce the energy required to move each bit, but the system still contains lasers, modulators, drivers, receivers and high-performance processors that generate heat. Co-locating these components concentrates thermal power in a small area. Temperature changes can alter the refractive index of photonic materials, shift the operating wavelength of resonant devices and reduce the stability of optical links. Lasers and detectors also have temperature-dependent performance. Effective cooling must therefore protect both the electronic circuits and the optical alignment, potentially requiring advanced heat spreaders, microfluidic cooling or new package materials capable of managing heat without disrupting optical operation.

Manufacturing and standardization will determine whether these technologies move beyond demonstrations and into large-scale computing infrastructure. Optical systems must be assembled with precise alignment, tested at high throughput and manufactured with acceptable yields. A defect in a photonic component or an imperfect fibre coupling point can reduce the performance of an entire package. The industry also needs common standards for optical interfaces, control protocols, chiplet communication, thermal specifications and testing procedures. Without interoperability, system designers may be locked into proprietary solutions, slowing adoption and increasing costs. Standardized interfaces could allow processors, photonic engines and memory devices from different suppliers to be combined in modular architectures.

The review presents optical compute interconnects not as a single replacement technology but as a broad architectural shift. The most successful systems are likely to combine electrical and optical links, assigning each connection type to the distance, bandwidth and latency requirements for which it is best suited. Electrical interconnects may continue to dominate very short local routes, while optical channels handle communication between chiplets, packages, racks or distant accelerator clusters. If engineers can reduce conversion losses, control heat, improve manufacturability and establish reliable standards, co-packaged optics could become a foundation for future AI and high-performance computing systems. The race to build faster artificial intelligence may therefore be decided not only inside the processor, but also in the invisible network of light carrying data between its many parts.

Subject of Research: Optical chip-to-chip interconnects and co-packaged optics for high-performance computing and artificial intelligence

Article Title: Co-packaged optics for high-performance computing and artificial intelligence

Article References: Kim, B., Choi, S.H., Zograf, G. et al. Co-packaged optics for high-performance computing and artificial intelligence. Nat Electron (2026). https://doi.org/10.1038/s41928-026-01681-6

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41928-026-01681-6

Keywords: Co-packaged optics, optical interconnects, photonic computing, chip-to-chip communication, artificial intelligence, high-performance computing, heterogeneous integration, 2.5D interposers, 3D stacking, data-centre technology

At the heart of CPO’s efficacy lies the optical waveguide, a fundamental component responsible for guiding light with minimal loss and distortion. Among various candidates, single-mode polymer waveguides have attracted significant attention due to their blend of mechanical flexibility, cost-effectiveness, and compatibility with existing electrical circuitry. These waveguides facilitate efficient coupling of laser light—especially from external sources—to PICs and enable seamless distribution of optical signals across the intricate architecture of CPO systems. Yet, despite these advantages, questions concerning their long-term stability, performance consistency, and resilience under demanding operational conditions have lingered.

Addressing these concerns, a team spearheaded by Dr. Satoshi Suda at Japan’s National Institute of Advanced Industrial Science and Technology (AIST) embarked on a comprehensive investigation into the stability and reliability of single-mode polymer waveguides fabricated on glass-epoxy substrates. Their recent experimental foray, documented in the IEEE Journal of Lightwave Technology, sheds illuminating insights on the potential of these organic waveguides to underpin next-generation optical interconnects in high-density photonic systems. The findings carry profound implications for the future design and deployment of optical components in co-packaged optics.

The fabrication process employed by Dr. Suda’s team centered on direct laser writing to sculpt 11-millimeter-long polymer waveguides onto FR4 glass-epoxy substrates. These substrates, known for their mechanical robustness and widespread use in printed circuit boards, offer an attractive platform for integrating optical structures alongside electronic components. Crucially, the waveguide cores were engineered with a cross-sectional dimension of approximately 9.0 micrometers by 7.0 micrometers, a size meticulously optimized to match the mode field diameter of standard single-mode optical fibers. This precise optimization ensures minimal coupling losses when interfacing with commercially available fiber optics, a vital requirement for scalable and practical CPO implementations.

One of the core performance metrics evaluated was polarization-dependent loss (PDL), which quantifies the differential attenuation of differently polarized light modes within the waveguides. Low PDL values, as demonstrated by the researchers, indicate that the waveguides can transmit optical signals with minimal polarization-induced distortion. Another crucial parameter, differential group delay (DGD), reflective of the timing disparity between different polarization components, was also found to be exceptionally low. These optical characteristics collectively contribute to robust, stable signal transmission by minimizing modal dispersion and preserving signal integrity over time and across different operational conditions.

Uniformity analysis across multiple fabricated samples further reinforced the waveguides’ reliability, revealing only slight variations in insertion loss and mode field dimensions. This consistency is pivotal for manufacturing scalability, as it ensures that components produced in large quantities will exhibit predictable and reproducible performance. In the realm of energy-efficient optical interconnects for CPO systems, such repeatability reduces the risk of bottlenecks or signal degradation that can undermine overall system throughput.

A particularly striking discovery pertained to the polarization extinction ratio (PER), a critical indicator of the waveguide’s ability to maintain a specific polarization state of light throughout its propagation. Within the experimental setting, the polymer waveguides exhibited PER values exceeding 20 decibels across all wavelengths defined by the coarse wavelength division multiplexing (CWDM4) standard—namely at 1271, 1291, 1311, and 1331 nanometers. This achievement aligns perfectly with the stringent Optical Interconnect Forum (OIF) specifications for ELS-based CPO systems, underscoring the waveguides’ suitability for commercial-grade optical communications where polarization fidelity is paramount.

Beyond optical performance alone, the study delved into operational stability under high optical power conditions. The team subjected the polymer waveguides on glass-epoxy substrates to continuous high-power illumination for up to six hours, an endurance test simulating real-world high-throughput data center environments. Impressively, the waveguides exhibited negligible degradation attributable to power-induced effects, maintaining stable insertion loss characteristics and mode profiles throughout the prolonged exposure. Complementing this resilience, minimal thermal heating was observed, affirming the capacity of these waveguides to sustain intensive operation without compromising physical integrity or optical function.

Of note, Furukawa Electric Co., Ltd. contributed the external laser sources utilized in these experiments, enabling stable high-power operation. This collaboration highlights the importance of synergy between optical component fabrication and reliable laser integration—a symbiosis critical for translating promising laboratory results into field-deployable CPO hardware. Dr. Suda articulates the broader ramifications, emphasizing that the demonstrated optical properties and robustness establish polymer waveguides as a viable technological building block for next-generation high-density co-packaged optical interconnects.

These revelations arrive at a pivotal moment when the insatiable demand for data processing and transmission bandwidth threatens to outstrip the capabilities of traditional copper-based interconnects. Photonic solutions, with their inherent advantages in speed and energy efficiency, are poised to revolutionize the landscape of data communications. Within this context, the validated performance and stability of single-mode polymer waveguides open new horizons for integrating photonics more deeply and reliably into electronic systems at scale.

Looking forward, the convergence of polymer-based waveguides with silicon photonics and advanced co-packaging techniques promises to unlock significant leaps in computing efficiency and network scalability. Ongoing research will doubtlessly explore refinements in polymer formulations, substrate engineering, and fabrication precision to further enhance waveguide performance and compatibility with diverse photonic and electronic elements. As these innovations progress from experimental validation to commercial application, the vision of seamlessly integrated, ultra-fast, and energy-efficient optical interconnects is rapidly becoming a technical and economic imperative.

In essence, the work led by Dr. Satoshi Suda and colleagues provides compelling evidence that polymer optical waveguides, when meticulously engineered and tested under rigorous conditions, can meet the demanding requirements of future co-packaged optical systems. Their optical clarity, polarization control, and high-power reliability compose a compelling narrative for the adoption of polymer photonics as a cornerstone technology supporting the next frontier of computing and telecommunications infrastructure worldwide.

Subject of Research: Not applicable

Article Title: High-Power Stability and Reliability of Polymer Optical Waveguide for Co-Packaged Optics

News Publication Date: 18-Feb-2025

Web References:
https://ieeexplore.ieee.org/abstract/document/10892053

References:
Satoshi Suda et al., IEEE Journal of Lightwave Technology, DOI: 10.1109/JLT.2025.3543339

Image Credits: Engineering at Cambridge at Openverse

Keywords

Optics, Applied optics, Electrical engineering, Nanotechnology, Photonics, Electromagnetism, Electronics, Waveguides