Coherent pluggable suits modern networks
Nov 07, 2025|
Coherent pluggable optics integrate high-capacity optical transmission directly into routers and switches, eliminating external transponders and streamlining network architectures. These compact modules support 100G to 800G data rates in small form factors like QSFP-DD and OSFP, making them essential for data center interconnects, metro networks, and IP-over-DWDM deployments.
The Economic Driver Behind Adoption
Network operators face mounting pressure to expand bandwidth while controlling costs. Traditional optical transport architectures require separate transponder equipment between routers and DWDM line systems, creating multiple conversion points that consume power, rack space, and capital. Coherent pluggable technology addresses this by consolidating optical and IP functions into a single device.
Bell Canada's deployment demonstrates the financial impact. The operator projects savings of $125 million CAD over ten years, driven by a 27% reduction in capital expenditures. Arelion achieved even more dramatic results with their 400G ultra-long-haul pluggables, cutting CAPEX by 35% and operational expenses by 84% when expanding network capacity. These aren't marginal improvements-they represent fundamental shifts in network economics.
The technology works by using digital signal processing and coherent detection within thumb-sized modules that plug directly into router ports. This eliminates not only the transponder hardware but also the associated cooling, power distribution, and management systems. In metro networks where distances range from 80 to 500 kilometers, coherent pluggable optics deliver the reach and capacity previously requiring dedicated optical transport equipment.
Power efficiency adds another economic dimension. Colt Technology Services reported 97% energy savings when deploying 800G ZR+ coherent pluggable optics compared to traditional architectures. As data centers face power constraints-with average rack densities climbing from 8 kW in 2022 to 17 kW in 2024 and projected to reach 30 kW by 2027-every watt saved translates directly to deployable capacity.
Technology Evolution: From 400ZR to 800ZR and Beyond
The Optical Internetworking Forum released the 400ZR implementation agreement in 2020, establishing interoperable standards for 400G coherent modules in QSFP-DD form factors. This standardization proved transformative. According to Cignal AI, 400ZR coherent optics achieved adoption rates three times faster than any previous coherent technology at similar maturity stages. By 2024, coherent pluggable accounted for all telecom bandwidth growth, with aggregate bandwidth from embedded optics actually declining year-over-year.
The 400ZR specification targets single-span links up to 120 kilometers, using QPSK modulation and concatenated forward error correction. For longer reaches, OpenZR+ extends capabilities to approximately 500 kilometers through enhanced FEC and flexible modulation schemes supporting 100G to 400G line rates. These modules maintain interoperability across vendors while accommodating diverse network requirements from point-to-point metro connections to multi-span ROADM networks.
Now the industry transitions to 800ZR. The OIF released the 800ZR implementation agreement in October 2024, doubling capacity while maintaining similar power and space requirements. Operating at 120 GBaud-twice the 60 GBaud used in 400ZR-these modules leverage 5-nanometer DSP technology instead of the 7nm process in earlier generations. This semiconductor advancement delivers 30-40% improvements in power efficiency per transmitted bit.
Marvell established an early lead in 800ZR DSP shipments, though the company released modules before the final 800ZR+ standard incorporated PCS (Physical Coding Sublayer) specifications for long-haul routes. Cisco/Acacia, Ciena, and component suppliers including Coherent and Lumentum have demonstrated 800ZR modules, with field trials completed across multiple service providers. Colt's enhanced 800G ZR+ trial doubled packet core capacity per link while reducing power consumption per bit by 33.3%.
The technology roadmap continues aggressively. The OIF began work on 1.6T coherent pluggable standards in 2024, with 1600ZR and 1600ZR+ implementation agreements in development. These ultra-high-capacity modules will serve both metro and long-haul applications, though they may require new form factors beyond current QSFP-DD and OSFP specifications to manage thermal and power requirements.
IP-Over-DWDM Architecture Transformation
The convergence of IP routing and optical transport represents more than incremental improvement-it fundamentally rearchitects network layers. Traditional networks maintain strict separation between packet processing in routers and wavelength management in optical systems. This bifurcation requires protocol conversions, separate management domains, and coordinated provisioning across teams with different tools and expertise.
IP-over-DWDM collapses these layers. Routers equipped with coherent pluggable directly generate DWDM wavelengths, enabling packets to traverse optical infrastructure without intermediate conversions. The architecture eliminates transponder shelves, OTN (Optical Transport Network) equipment, and the grey optics typically connecting routers to transport systems. Network operators can provision services through router interfaces alone, treating wavelengths as extended Ethernet links.
This convergence introduces operational challenges, particularly in service provider networks with established organizational structures. A Heavy Reading survey found that 39% of communications service providers favor optical controllers to manage coherent pluggable in routers, while 22% prefer IP controllers and 20% support hierarchical approaches. Notably, 16% remained undecided despite years of evaluation-organizational alignment, not just technology choice, determines successful implementation.
The management complexity stems from conflicting requirements. IP teams prioritize dynamic routing, automated failover, and application-layer services. Optical teams focus on wavelength planning, dispersion management, and physical layer optimization. When coherent pluggable physically reside in routers but require optical expertise for link engineering, responsibility boundaries blur. Some operators resolve this through automation platforms spanning both domains, using standardized YANG models and NETCONF protocols to abstract complexity.
Disaggregated architectures amplify these benefits. Open line systems allow operators to insert wavelengths from router-based coherent pluggable rather than relying exclusively on transponders from the same vendor as the ROADM equipment. Approximately 70% of networks using router-based coherent pluggable deploy over open line systems, according to industry data. This multi-vendor approach diversifies supply chains and accelerates innovation, though it demands rigorous interoperability testing and comprehensive link budgets.
Optical bypass strategies further optimize costs. Rather than routing transit traffic through IP routers at every node-consuming power for both the pluggable and the forwarding engine-wavelengths pass through ROADMs optically. This approach proves most effective in linear or ring topologies with moderate mesh complexity. For highly interconnected networks, point-to-point pluggable links may offer simpler operations despite higher transceiver counts.
Bandwidth Demands Driven by AI Workloads
Data center bandwidth purchasing surged 330% between 2020 and 2024, with AI workloads as the primary catalyst. This explosive growth differs fundamentally from previous traffic increases. Traditional cloud applications generate north-south traffic patterns-data moving between end users and servers. AI training creates massive east-west flows as GPUs exchange gradients and model parameters across thousands of nodes within and between data centers.
The scale is staggering. Modern AI training clusters require 400 Gbps to 1.6 Tbps connections between nodes, with latency thresholds measured in microseconds. A single Large Language Model training run can generate petabytes of data movement. As training becomes distributed across multiple facilities-81% of data center operators expect this trend according to recent surveys-the pressure on data center interconnect infrastructure intensifies dramatically.
Metro dark fiber purchases rose 268% from 2023 to 2024, while long-haul dark fiber grew 53% in the same period. Geographic patterns reveal AI's infrastructure impact. Memphis, Tennessee saw long-haul and metro bandwidth demand explode from 0.3 terabits in 2023 to 13.2 terabits in 2024-a 4,300% increase driven by hyperscaler land and power acquisitions. Salt Lake City experienced 348% growth for similar reasons.
Coherent pluggable technology directly addresses these AI networking requirements. The modules' high capacity and efficient scaling align with AI's voracious bandwidth appetite. Their integration into routers simplifies the massive parallel connections AI clusters demand. Power efficiency becomes critical-98% of data center operators cite pluggable optics as important for reducing power consumption and physical footprint, according to a 2025 survey of 1,300 global data center decision-makers.
Projection models indicate bandwidth needs will continue accelerating. Data center experts predict a minimum sixfold increase in DCI bandwidth over the next five years-representing 40-60% compound annual growth, more than double typical historical rates. To support this demand, 87% of operations expect to need 800 Gbps or faster wavelengths for data center interconnects by 2030, with 43% of new data center construction dedicated specifically to AI workloads.
The coherent pluggable market responds accordingly. Cignal AI data shows 400G pluggable shipments continuing with long-tail growth through 2027, while 800ZR and 1600ZR deployments ramp simultaneously. After 2026, telecom bandwidth growth will be overwhelmingly dominated by pluggable rather than embedded solutions as these modules extend from metro into long-haul networks supporting distributed AI training infrastructure.
Form Factor Competition and Technical Trade-offs
Two primary form factors compete in the coherent pluggable market: QSFP-DD and OSFP. QSFP-DD currently dominates shipments due to alignment with widely adopted host platform slots in existing routers and switches. The smaller footprint enables higher port densities-critical for maximizing faceplate real estate in chassis-based systems. Most 400ZR implementations use QSFP-DD, establishing an installed base that influences upgrade paths.
OSFP offers advantages for higher-speed and higher-power applications. The larger form factor provides superior thermal management and power delivery, supporting the demands of 800G and future 1.6T modules. Some vendors offer both form factors at 800ZR, allowing network operators to choose based on their specific density versus thermal requirements. OSFP's power budget accommodates high-transmit-power variants needed for legacy ROADM architectures or extended unamplified reaches.
Technical specifications reveal critical performance parameters. Standard 400ZR modules transmit at -10 dBm launch power and receive down to -21 dBm, supporting 80-120 km single-span links. High-transmit (HT) variants boost launch power to 0 dBm or +1 dBm, extending reach in ROADM networks or enabling longer unamplified point-to-point connections. These enhanced modules incorporate tunable optical filters (TOF) to minimize adjacent channel interference in colorless ROADM architectures.
Distance capabilities segment the market. Standard ZR addresses metro applications to 120 km. ZR+ extends reach to approximately 500 km through stronger FEC and flexible modulation, serving regional networks. Ultra-long-haul (ULH) pluggables push distances beyond 2,000 km with amplification, competing directly with embedded transponders in long-haul segments. Arelion's successful field trial demonstrated 400G ULH transmission over 2,253 kilometers at 112.5 GHz spectrum with healthy margins.
Modulation formats adapt to distance-capacity trade-offs. QPSK provides maximum reach at lower spectral efficiency. 16-QAM increases capacity for moderate distances. Higher-order modulation schemes like 64-QAM maximize throughput on short, high-quality links. Advanced modules support programmable modulation, allowing operators to optimize for specific route characteristics and traffic requirements.
Forward error correction represents another critical dimension. 400ZR uses concatenated FEC with approximately 15% overhead. OpenZR+ employs o-FEC (open FEC) with higher correction capability, enabling longer reaches and operation over more challenging optical paths. The stronger FEC comes at cost-increased latency from additional processing and higher power consumption. Network operators balance these factors based on application priorities.
Standards, Interoperability, and Ecosystem Development
Open standards drive coherent pluggable adoption by enabling multi-vendor ecosystems and preventing lock-in. The OIF's 400ZR implementation agreement established baseline specifications for optical characteristics, Ethernet client mappings, frame formats, and FEC. This foundational work created genuine interoperability-operators can mix pluggables from different suppliers with confidence in basic functionality.
The OpenZR+ MSA expanded capabilities beyond OIF 400ZR's scope. Published specifications cover extended reach, flexible line rates from 100G to 400G, and support for OTN client mappings. These enhancements address service provider requirements for metro ROADM networks and mixed traffic types. The MSA approach complements OIF's formal standardization, providing faster iteration on emerging requirements while maintaining commitment to interoperability.
Regular interoperability demonstrations validate standards compliance. OIF-organized plugfests bring together equipment vendors, module suppliers, and operators to test cross-vendor combinations. Successful demonstrations at OFC 2024 and ECOC 2024 showcased 800ZR interoperability across multiple providers, proving the technology ready for production deployment. These events identify edge cases and drive refinement of specifications before widespread adoption.
The Common Management Interface Specification (CMIS) addresses operational integration challenges. CMIS defines standardized management interfaces for coherent modules, enabling consistent monitoring and control regardless of vendor. Support for CMIS in coherent pluggable allows operators to retrieve performance metrics, adjust operational parameters, and coordinate line-side and host-side states through uniform APIs. Version 5.2, released in 2024, adds enhancements specifically for coherent applications including support for C+L band operation.
OpenROADM specifications contribute another layer of standardization. The OpenROADM MSA defines optical specifications and APIs for creating multi-vendor ROADM networks. Coherent pluggable supporting OpenROADM modes can interoperate with disaggregated line systems from different manufacturers, expanding deployment flexibility. Some advanced modules support both OpenZR+ and OpenROADM modes, allowing operators to select appropriate profiles based on network segment requirements.
Industry collaboration extends beyond technical specifications. The Telecom Infra Project's MANTRA subgroup published architectural guidelines for IPoDWDM deployments, addressing real-world integration challenges. Proof-of-concept activities bring together operators including Vodafone, Telefonica, Orange, and Deutsche Telekom with equipment and component vendors to validate designs. These collaborative efforts accelerate deployment by de-risking implementations and documenting best practices.
Standards work continues evolving. The OIF's efforts on 1600ZR specifications will establish baselines for next-generation modules. Questions remain about required power budgets, optimal form factors, and cooling approaches at these higher speeds. Early standardization enables component suppliers to align development roadmaps, reducing fragmentation and accelerating ecosystem maturity when products reach market.
Network Operator Adoption Patterns
Service provider adoption diverges from hyperscaler patterns in important ways. Hyperscalers pioneered coherent pluggable for metro data center interconnect, focusing on point-to-point links with homogeneous equipment and centralized control. Their networks feature spine-and-leaf architectures with standardized distances and traffic patterns. This environment suits 400ZR's initial specifications perfectly-single-span links to 120 km carrying Ethernet at scale.
Communications service providers operate more diverse networks. They manage existing ROADM infrastructure with multiple vendors, support varied service types including OTN and private line, and maintain separate IP and optical organizations. A Heavy Reading survey found 65% of CSPs believe coherent pluggable will require OTN OAM (Operations, Administration, and Maintenance) functions for transport applications. Only 16% consider ZR+ sufficient for all use cases, with 45% stating that mesh ROADM applications specifically need OTN support.
This complexity influences deployment strategies. Tier 1 operators prioritize manageability and point-to-multipoint functionality significantly higher than smaller providers. Managing thousands of coherent modules distributed across router faceplates in multiple cities demands sophisticated automation. Heavy Reading's 2024 survey revealed manageability topped the priority list when price and power consumption were excluded as factors, selected by 50% of global respondents.
Spine-and-leaf architectures generate interest in service provider WANs despite their origins in data center networks. The 2025 Heavy Reading survey found 54% of CSPs considering spine-and-leaf for WAN deployment, while 26% already use the architecture-surprising given its newness in telecom. Spine-and-leaf offers predictable performance, simplified path management, and natural alignment with IP-over-DWDM by treating optical layers as extended switch fabrics. However, it represents a fundamental departure from traditional telecom ring and mesh designs.
Deployment timelines reflect this operational complexity. While hyperscalers moved to production quickly, CSPs proceed more deliberately. Trials validate performance over existing fiber plants, test integration with incumbent management systems, and verify interoperability across vendor combinations. Bell Canada's multi-year deployment demonstrates the careful approach-projecting decade-long savings rather than immediate transformation. The conservative pace reflects prudent risk management in networks carrying diverse customer services with strict SLAs.
Geographic variations add another dimension. North American operators lead in coherent pluggable adoption, driven by hyperscaler demand and progressive regulatory environments. European service providers follow closely, motivated by competitive pressures and sustainability mandates. Asian markets show mixed patterns-Singapore and South Korea deploy aggressively while other regions move more cautiously. China's unique ecosystem favors domestic suppliers with integrated equipment and optics, creating less opportunity for standalone coherent pluggable from pure component vendors.
Enterprise networks represent emerging adoption territory. Large enterprises with distributed data center estates evaluate coherent pluggable for private interconnects. Healthcare systems, financial institutions, and research networks investigate the technology as costs decline and operational simplicity improves. The total addressable market expands as 100G coherent modules in QSFP28 form factors target edge applications previously served by grey or CWDM optics.
Competing with Embedded Solutions
Coherent pluggable doesn't replace embedded transponders entirely-both technologies serve complementary roles. Ciena's WaveLogic 6 Extreme, a 1.6T embedded coherent modem, added 20 customers in a single fiscal quarter after commercial availability. Applications requiring maximum spectral efficiency or ultra-high performance still favor embedded solutions, particularly in submarine, long-haul, and capacity-constrained routes.
The fundamental trade-off involves optimization priorities. Pluggable optimize for space, power, and integration with host routers. They sacrifice some performance margin to meet small form factor and thermal constraints. Embedded solutions optimize for raw capacity, spectral efficiency, and link margin. Built on dedicated line cards with superior cooling and power delivery, they push closer to Shannon limits and extract maximum bits per hertz from deployed fiber.
Cost considerations vary by application. For metro links where a single pluggable in a router port suffices, total cost strongly favors the pluggable approach-no separate chassis, no independent power and cooling, no grey client optics. For long-haul routes requiring multiple ROADM hops and sophisticated spectrum management, embedded transponders may deliver better economics through tighter channel spacing and superior spectral efficiency. The crossover point shifts as pluggable technology advances and volumes drive down pricing.
Market data shows the coexistence clearly. According to Cignal AI, both 1.2T+ embedded solutions and 400G/800G pluggable contribute to bandwidth growth in 2025 and beyond. Each technology addresses distinct requirements. Network operators increasingly evaluate applications individually rather than adopting blanket policies. A service provider might deploy pluggable for metro aggregation and business services while using embedded modules for intercity trunks and international routes.
Technology roadmaps suggest converging capabilities over time. As coherent pluggable DSPs migrate to 3-nanometer process nodes and achieve higher baud rates, performance gaps narrow. Conversely, embedded solutions adopt techniques from pluggable including co-packaged optics and advanced digital subcarriers. The boundary between categories blurs, with some solutions offering pluggable form factors but performance approaching embedded levels by accepting higher power budgets.
Challenges and Operational Considerations
Link engineering complexity increases with coherent pluggable deployments. Unlike traditional transponders where vendor-supplied engineering tools calculate feasible routes, operator-managed pluggable require direct expertise in optical physics. Parameters including chromatic dispersion, polarization mode dispersion, OSNR (Optical Signal-to-Noise Ratio), and nonlinear effects must be carefully budgeted. Mistakes lead to marginal links that fail under stress or require expensive remediation.
This challenge proves particularly acute in ROADM networks. Add/drop losses, filtering effects, and wavelength-dependent insertion losses create complex optical environments. High-transmit-power pluggable help overcome some limitations but introduce adjacent channel interference risks in colorless architectures. Operators need sophisticated planning tools and photonic layer expertise-capabilities traditionally concentrated in optical engineering teams rather than IP network operations.
Software management fragmentation remains problematic. Router vendors, pluggable module suppliers, and ROADM equipment manufacturers each provide management interfaces with different capabilities and abstractions. Achieving unified visibility across all three requires integration work and often custom development. While CMIS and OpenConfig provide standardized foundations, implementation variations and vendor-specific extensions complicate deployment.
Temperature sensitivity affects coherent pluggable performance. The tight integration with routers means modules experience thermal environments determined by chassis cooling rather than dedicated optical equipment. High ambient temperatures or inadequate airflow can degrade link margins or trigger thermal shutdowns. Data center operators must account for optical layer requirements when designing cooling strategies-consideration traditionally unnecessary when optics resided in separate bays with independent thermal management.
Power consumption at scale deserves careful attention. While individual coherent pluggable consume less power than transponders, dense router faceplates with 32 or 64 ports can draw substantial current. A fully populated chassis might require several kilowatts just for optics, separate from router forwarding and linecard power. The consolidated power draw stresses chassis power supplies and raises cooling requirements. Network operators must validate thermal and power specifications across full-scale deployments, not just prototype configurations.
Upgrade cycles create coordination challenges. IP router refresh typically occurs on 3-5 year cycles, while optical infrastructure operates for 7-10 years or longer. When routers reach end-of-life, operators face decisions about retained coherent pluggable. Can modules be reused in new platforms? Do they support emerging software interfaces? These lifecycle mismatches complicate planning and may strand investment earlier than traditional optical equipment.
Supply chain diversity presents both opportunities and risks. Multiple sources for coherent pluggable reduce dependence on single vendors and improve negotiating leverage. However, qualifying multiple vendors requires extensive testing, and mixing sources in production networks demands careful management of firmware versions and feature sets. Some operators standardize on primary and secondary vendors to balance diversity against operational complexity.
The Path Forward
Network architecture evolution continues accelerating, driven by insatiable bandwidth demands and economic pressures. Coherent pluggable technology proves central to this transformation, enabling convergence of IP and optical layers while delivering compelling cost and efficiency benefits. The momentum established by 400ZR carries forward into 800ZR and future terabit-scale implementations.
Several developments will shape near-term progress. The 800ZR ecosystem matures through 2025 as additional vendors deliver products and field deployments expand beyond early adopters. Standards work on 1600ZR lays groundwork for the next capacity jump, though thermal and power challenges may push those implementations to later in the decade. Meanwhile, 100G coherent pluggable in QSFP28 form factors target edge and access networks, extending the technology's reach to new applications.
Management and automation improvements reduce operational friction. Enhanced software platforms spanning IP and optical domains streamline provisioning and monitoring. Machine learning algorithms optimize modulation and FEC parameters dynamically based on real-time link conditions. Closed-loop automation handles common operations without human intervention, reducing skill requirements and accelerating service delivery.
The distributed AI training trend amplifies coherent pluggable importance. As Large Language Models grow too large for single-site training, interconnecting GPU clusters across metro and regional distances becomes critical. Coherent pluggable provide the capacity, latency, and economic efficiency this application demands. Network operators positioning themselves to serve AI infrastructure will find coherent pluggable capabilities increasingly central to their competitive position.
Open disaggregated architectures gain traction as operators prioritize flexibility and supply chain resilience. The success of open line systems and interoperable coherent pluggable demonstrates viability of multi-vendor approaches. Further disaggregation extending to white-box routers and switches intensifies these trends, reshaping traditional equipment and optics markets. Component suppliers and software providers capture value previously concentrated in integrated systems.
Sustainability considerations influence technology choices as regulatory pressure and customer requirements emphasize carbon reduction. The power efficiency of coherent pluggable-particularly newer generations on advanced process nodes-aligns with these mandates. Network operators can increase capacity while stabilizing or reducing power consumption, achieving both business and environmental objectives. This dual benefit strengthens the technology's position in long-term infrastructure planning.
Frequently Asked Questions
What's the difference between 400ZR and 400ZR+?
400ZR supports single-span links up to 120 km using QPSK modulation and concatenated FEC, optimized for data center interconnect. 400ZR+ extends reach to approximately 500 km through enhanced open FEC and supports flexible modulation and multiple line rates from 100G to 400G. ZR+ modules can operate in ROADM networks and support OTN client mappings, addressing service provider requirements beyond hyperscaler metro applications.
Can coherent pluggable from different vendors interoperate?
Yes, when compliant with OIF or OpenZR+ specifications. Standardized optical characteristics, modulation formats, FEC schemes, and framing enable multi-vendor interoperability. However, advanced features beyond baseline specifications may vary between suppliers. Operators should validate specific vendor combinations in their network environments, particularly for ROADM deployments with multiple amplifier spans. Regular industry plugfests demonstrate cross-vendor compatibility across expanding feature sets.
How do coherent pluggable handle network management?
Modern coherent pluggable support CMIS for standardized monitoring and control. They report performance metrics including pre-FEC bit error rates, OSNR estimates, chromatic dispersion, and temperature through common interfaces. Advanced modules implement C-CMIS extensions for coherent-specific parameters. Integration with router management systems allows unified visibility, though achieving full operational convergence between IP and optical layers requires sophisticated software platforms that span both domains.
What distances can 800ZR modules achieve?
Standard 800ZR targets 80-120 km single-span links similar to 400ZR. Enhanced 800ZR+ extends reach to 500+ km through stronger FEC and optimized modulation. Ultra-long-haul variants under development aim for 1,000-2,000 km with amplification. Actual distances depend on fiber quality, ROADM insertion losses, and required margin. Higher transmit power variants (+1 dBm) extend reach in both unamplified and amplified configurations by improving link budgets.
Do coherent pluggable work with existing DWDM equipment?
Compatibility depends on the optical line system. Coherent pluggable operate over open line systems supporting alien wavelengths without issues when properly engineered. Legacy ROADM architectures may require high-transmit-power modules to compensate for insertion losses and filtering effects. Some older systems lack sufficient channel bandwidth or introduce excessive polarization-dependent loss. Operators should conduct detailed link engineering including dispersion budgets and nonlinear calculations before deploying pluggable in existing infrastructure.
How does power consumption compare to traditional transponders?
Individual coherent pluggable consume less power than dedicated transponders-400ZR modules typically draw 12-15W versus 100-150W for line-card-based transponders. However, at scale with multiple ports, total power per chassis can be substantial. The key advantage comes from eliminating separate grey client optics, DWDM transponder shelves, and associated cooling infrastructure. System-level power savings of 64-97% have been reported by operators deploying converged IP-optical architectures with coherent pluggable.
References
Cignal AI Transport Hardware Report, 2024-2025
Heavy Reading Coherent Optics Survey, 2024-2025
OIF 400ZR and 800ZR Implementation Agreements
Ciena Global Data Center Networking Report, 2024-2025
Acacia Communications field trial results, 2024
Light Reading industry analysis, 2023-2025
Dell'Oro Group optical market forecasts
Zayo Bandwidth Report, 2024