![]() |
市場調查報告書
商品編碼
2109219
2026 年至 2035 年 AI 晶片光 I/O 市場的機會、成長促進因素、產業趨勢分析與預測。Optical I/O for AI Chips Market Opportunity, Growth Drivers, Industry Trend Analysis, and Forecast 2026 - 2035 |
||||||
全球人工智慧晶片光 I/O 市場預計到 2025 年將達到 11 億美元,並以 27.8% 的複合年成長率成長,到 2035 年將達到 140 億美元。

人工智慧 (AI) 運算基礎設施的快速擴張,催生了對能夠提供更高頻寬、更低延遲和更高能效的先進光連接解決方案的強勁需求。共封裝光元件 (CPO) 的日益普及、政府旨在加強國內半導體製造業的支持以及對矽光電技術的持續投入,都在加速市場成長。高速光纖網路標準的日益普及,進一步推動了向下一代 AI 硬體架構的轉型。基於矽光電的光子積體電路和光收發器的持續創新,正在提升先進運算環境的可擴展性並降低功耗。隨著 AI 工作負載日益複雜且資料密集,光 I/O 技術正成為提升互連性能和系統效率的關鍵元件。對高效能運算基礎設施日益成長的需求,以及對半導體和光電系統的持續投入,預示著全球 AI 晶片光 I/O 市場將在長期內顯著擴張。
| 市場範圍 | |
|---|---|
| 開始年份 | 2025 |
| 預測期 | 2026-2035 |
| 初始市場規模 | 11億美元 |
| 預計金額 | 140億美元 |
| 複合年成長率 | 27.8% |
受人工智慧資料中心快速擴張以及對頻寬更高、功耗更低的互連技術(相較於傳統電連接)日益成長的需求驅動,人工智慧晶片的光I/O市場呈現強勁成長動能。大規模人工智慧運算基礎設施投資的增加,正加速滿足先進處理需求的光互連解決方案的普及。同時,公共和私人對半導體製造和矽光電生產的投資不斷增加,正在強化國內供應鏈,並推動下一代光技術的商業化,從而進一步促進市場的持續成長。
預計到2025年,線性可插拔光元件(LPO)將成為主要產品類型,佔63.2%的市佔率。該領域持續受到廣泛認可,原因在於其易於與現有資料中心基礎設施整合,相比傳統插入式解決方案部署複雜度更低,並且能夠經濟高效地過渡到共封裝光元件。 LPO技術在保持低功耗的同時,也能提供極高的頻寬,因此非常適合要求嚴苛的人工智慧工作負載和高階運算應用。
預計到2025年,矽光電(SiPh)市場規模將達到6.113億美元。該領域因其與CMOS製造流程的兼容性、卓越的整合能力以及為大規模生產提供高頻寬、高能效光連接模組的能力而得到廣泛應用。人工智慧運算平台、超大規模資料中心和高效能運算環境等領域的日益普及推動了市場需求的持續成長,進一步促進了對矽光電技術的投資。
到2025年,北美人工智慧晶片的光I/O市佔率將達到57.8%。這一成長得益於成熟的半導體生態系統、對人工智慧運算基礎設施的大量投資,以及領先的人工智慧硬體、光纖網路技術和矽光電解決方案開發商的存在。對高速、低延遲光連接模組的需求不斷成長,以滿足日益複雜的人工智慧訓練和推理工作負載,這進一步鞏固了該地區的主導地位。
The Global Optical I/O for AI Chips Market was valued at USD 1.1 billion in 2025 and is estimated to grow at a CAGR of 27.8% to reach USD 14 billion by 2035.

Rapid expansion of artificial intelligence computing infrastructure is creating strong demand for advanced optical connectivity solutions capable of delivering greater bandwidth, lower latency, and improved energy efficiency. The increasing deployment of co-packaged optics (CPO), supportive government initiatives aimed at strengthening domestic semiconductor manufacturing, and growing investments in silicon photonics technologies are accelerating market growth. Rising adoption of high-speed optical networking standards is further supporting the transition toward next-generation AI hardware architectures. Continuous innovation in silicon photonics-based photonic integrated circuits and optical transceivers is enabling higher scalability while reducing power consumption across advanced computing environments. As AI workloads become increasingly complex and data-intensive, optical I/O technologies are emerging as a critical component for improving interconnect performance and system efficiency. The growing need for high-performance computing infrastructure, combined with sustained investment in semiconductor and photonics ecosystems, is expected to drive long-term expansion of the global optical I/O for AI chips market.
| Market Scope | |
|---|---|
| Start Year | 2025 |
| Forecast Year | 2026-2035 |
| Start Value | $1.1 Billion |
| Forecast Value | $14 Billion |
| CAGR | 27.8% |
The optical I/O for AI chips market is experiencing strong momentum as the rapid expansion of AI data centers increases demand for interconnect technologies that provide significantly higher bandwidth and lower power consumption than conventional electrical connections. Growing investments in large-scale AI computing infrastructure are accelerating the adoption of optical interconnect solutions capable of supporting advanced processing requirements. At the same time, increasing public and private investment in semiconductor manufacturing and silicon photonics production is strengthening domestic supply chains and supporting the commercialization of next-generation optical technologies, further contributing to sustained market growth.
The linear pluggable optics (LPO) segment accounted for a 63.2% share in 2025, making it the leading product category. The segment continues to gain widespread acceptance because of its straightforward integration with existing data center infrastructure and its cost-efficient migration path toward co-packaged optics with fewer deployment complexities than conventional pluggable solutions. LPO technology delivers exceptionally high bandwidth while maintaining low power consumption, making it well suited for demanding AI workloads and advanced computing applications.
The silicon photonics (SiPh) segment reached USD 611.3 million in 2025. The segment has achieved strong market adoption due to its compatibility with CMOS manufacturing processes, superior integration capabilities, and ability to deliver high-bandwidth, energy-efficient optical interconnects at large production volumes. Expanding deployment across AI computing platforms, hyperscale data centers, and high-performance computing environments continues to strengthen demand and encourage greater investment in silicon photonics technologies.
North America Optical I/O for AI Chips Market accounted for 57.8% share in 2025. Regional growth is supported by a well-established semiconductor ecosystem, significant investments in AI computing infrastructure, and the presence of leading developers of AI hardware, optical networking technologies, and silicon photonics solutions. Rising demand for high-speed, low-latency optical interconnects designed to support increasingly complex AI training and inference workloads continues to reinforce the region's leadership position.
Key players operating in the global optical I/O for AI chips market include Ayar Labs, Broadcom Inc., Celestial AI, Ciena Corporation, Coherent Corp., Furukawa Electric Co., Hamamatsu Photonics, Intel Corporation, Lightmatter, Lumentum Holdings Inc., MACOM Technology Solutions, Marvell Technology Group, NewPhotonics, Nexus Photonics, PhotonicX AI Pte. Ltd., POET Technologies Inc., Ranovus Inc., Salience Labs, Scintil Photonics, and Sumitomo Electric Industries. Companies operating in the optical I/O for AI chips market are strengthening their market position by investing heavily in silicon photonics innovation, advanced optical interconnect technologies, and next-generation packaging solutions. Many organizations are expanding research and development activities to improve bandwidth, reduce power consumption, and enhance scalability for AI-driven computing environments. Strategic collaborations with semiconductor manufacturers, cloud infrastructure providers, and technology developers are helping companies accelerate commercialization and broaden product adoption. Businesses are also increasing manufacturing capabilities, strengthening supply chain resilience, and introducing integrated optical solutions tailored to evolving AI infrastructure requirements.