Available in December, 2025
Co-packaged optics (CPO) addresses bandwidth and latency bottlenecks by moving optical interfaces close to the electrical die, shortening high-speed electrical paths and enabling denser, low-power optical fabrics. This results in higher aggregate bandwidth per accelerator and low energy/bit — thereby aligning with the economics and engineering needs of AI clusters, which is where CPO is currenly being positioned . CPO technology is still immature at the present time and the forecasts contained in this report paint different standards and product scenarios and how they will impact the the evolving CPO market
In 2020 CIR was the first analysis firm to publish a co-packaged optics (CPO) market report. We now think that by 2026, CPO will become a major interconnect technology in AI data centers throughout the world. In this report, we build a roadmap for the transition to CPO and show where business value will be created. Topics covered by this report include:
- How CPO products, technology and standards will evolve in the future
- CPO transition guidance for AI data center managers including case studies and profiles of CPO current trials
- Roadmaps of 30+ key CPO firms and how they identify their roadmaps with AI and rebrand CPO as primarily an AI platform. Also, a review of the most exciting CPO startups and where these firms see their competitive advantage
- CPO for on-rack, rack-to-rack and AI data center interconnection and the products that these applications inspire
A primary goal for this CIR report is to update CPO forecasts with breakouts by application, speed and technology, network segment and type of data center. Our new projections take into consideration how the rise of AI has impacted CPO technology and its market dynamics. CPO now promises a practical path to scaling multi-rack AI clusters. Meanwhile, major AI vendors, such as AI-accelerator suppliers and system integrators, are designing CPO solutions and positioning them for 2025–2026 deployments in AI data centers.
Table of Contents
Executive Summary
E.1 Market environment and situations influencing the market
E.2 Likely technical and market evolution of the CPO market in the AI sector
E.3 Emerging opportunities for CPO connectivity in the AI data center
E.3 Summary of CPO forecasts, 2026-2035
Chapter One: Introduction
1.1 Background to this report
1.2 Goals and scope of this Report
1.3 Plan of this report
Chapter Two: Co-packaged Optics: A Technology, Components and a Business Opportunity Evolving
2.1 CPO in traditional AI data centers
2.2 CPO for next-gen AI data centers
2.2.1 CPO data rates
2.2.2 Other technical issues that will impact CPO product design
2.3 Thermal management
2.4 Electrical interfaces and channel Loss
2.4.1Optical interfaces
2.4.2 Integration, manufacturing, CPO and test equipment
2.5 Optical engines
2.6 Fiber, Connector and Couplers for CPO
2.6.1 Thoughts from Broadcom and NVIDIA
2.6.2 Vertical coupling
2.7 Fiber solutions for CPO
2.7.1 Multi-core fibers
2.7.2 Reducing the pitch
2.8 Lasers for CPO
2.8.1 External lasers: Advantages and disadvantages
2.8.2 ELSFP and its coming use
2.8.3 Integration strategies
Chapter Three Emerging CPO Standards and their Implications
3.1 The emergence of CPO and the OIF
3.1.1 OIF Co-Packaging Framework
3.1.2 OIF standards for 1.6T and 3.2T CPO Module
3.1.3 External Laser Small Form Pluggable ELSFP IA
3.1.4 Telemetry and management
3.1.4 OIF’s CEI-112G XSR / XSR+ PAM4
3.2 Advanced Photonics Coalition
3.3 CPO and Ultra Ethernet
3.4 UCIe standard and its relationship to CPO
3.5 The CPO standards process in China
Chapter Four CPO Applications in AI Data Centers
4.1 AI/ML service, bandwidth demand and low latency
4.2 CPO in data center interconnection
4.3 Hyperscalers
4.3.1 A suitable case for CPO: Types of switches used
4.3.2 The CPO future for the hyperscalers
4.3.3 The use of NPO in hyperscale data centers
4.4 CPO in smaller data centers including co-location and enterprise data centers
4.5 CPO and edge data centers
4.6 CPO and high-performance computing centers
Chapter Five Today’s Real World CPO Deployment Problems, their Solutions and Alternatives
5.1 Ecosystem disruption
5.2 Operational complexity
5.3 Managing a CPO deployment plan
5.4 Technology choices
5.5 Full-scale implementation
5.6 Thermal management
5.7 Recent CPO Trials, demos and takeaways
5.8 Smooth transition to CPO #1: Linear Pluggable Optics
5.8.1 Latency and bandwidth
5.8.2 DSP and cost strategy
5.8.3 Power efficiency and electrical budgets
5.8.4 Thermal and mechanical simplicity
5.8.5 Form factors
5.9 Smooth transition to CPO #2: Near-pluggable optics (NPO)
5.9.1 Latency and bandwidth
5.9.2 DSP strategy, cost strategy and DSP vendor preferences
5.9.3 Power efficiency
5.9.4 Form factors and need for redesign
Chapter Six: Ten-Year Forecasts for AI Data Center CPO Markets
6.1 Forecasting methodology and assumptions
6.1.1 Patterns of network adoption
6.1.2 A Note on pricing
6.2 CPO by type of AI data centers (hyperscale, colocation, edge, enterprise and other)
6.2.1 Forecast by data rate and type of AI data center
6.2.2 Forecast by network location and type of AI data center
6.3 CPO in enterprise AI data centers: A forecast
6.4 CPO in HPC data centers: A forecast
6.5 Forecast of influential CPO components
Chapter Seven: Profiles: Suppliers and Influencers
7.1 AMD (United States)
7.1.1 Acquisition of Enosemi
7.1.2 AMD and Ranovus
7.2 Ayar Labs (United States)
7.2.1 Product Evolution
7.2.2 Alchip Technologies/TSMC alliance
7.2.3 Partnership with Quantifi Photonics
7.2.4 Partnership with HPE
7.2.5 Ayar the AI Infrastructure
7.2.6 Financing and collaborators
7.3 Broadcom (United States)
7.3.1 Emerging CPO product range
7.3.2 The Bailley switch
7.3.3 Alliance with Tencent
7.3.4 Long-term CPO strategy at Broadcom
7.4 Ciena/Nubis
7.4.1 Data Center Interconnects
7.4.2 Ciena’s Long-term Participation in the CPO Market
7.5 Cisco
7.6 Coherent
7.7 Corning
7.8 Furukawa Electric
7.9 Hisense Broadband
7.10 Huawei
7.11 IBM
7.11.1 History in CPO
7.11.2 Future Plans for CPO
7.12 Intel
7.12.2 Integrated optics programs
7.12.1 Optical Compute Interconnect
7.13 Kyocera
7.14 Lightmatter
7.15 Lumentum
7.16 Marvell
7.16.1 TERalynx Switch and current technology
7.16.2 Future Plans for CPO
7.17 Meta
7.19 Molex
7.19.1 Molex unveiled its External Laser Source Interconnect System (ELSIS
7.21 NVIDIA
7.21.1 External lasers
7.22.2 CPO in the NVIDIA switch family
7.22.3 CPO in the NVIDIA interconnect and network family
7.22.4 The future of CPO at NVIDIA
7.22 POET Technologies
7.23 Optical interposer
7.24 PICS
7.25 Evolution of technology
7.26 Quantifi
7.27 Ranovus
7.27.1 CPO architecture
7.28 SENKO
7.29 Skorpios
7.30 Sumitomo Electric
7.31 TE Connectivity
7.31 Teramount
7.32 Emerging start ups
7.33 Third-party suppliers and systems integrators
About the Author
