UTStarcom Unveils Optical Circuit Switching (OCS) Solution for AI Data Centers at CIOE 2026

UTStarcom debuts Silicon Photonics and MEMS-based OCS concept platforms for AI data center networking at CIOE 2026.

Rhea-AI Summary

UTStarcom (UTSI) showcased concept prototypes of its next-generation Optical Circuit Switching (OCS) platform for AI data center networks at CIOE 2026 in Shenzhen.

The company presented two 64x64-radix OCS prototypes: UOS64, based on Silicon Photonics photonic integrated circuits, and MOS64, using MEMS micro-mirror technology. Both are designed to create a dynamic, all-optical switching layer that offloads high-volume AI traffic, reduces latency and packet loss, and lowers power, cooling, and transceiver requirements.

UTStarcom also debuted its SONiC-based OCS software and SOO Station management platform, enabling automated optical-path provisioning and topology-on-demand via SDN and open interfaces.

Market Context

On Mar 24, UTStarcom disclosed OCS prototype development targeted for the second half of 2026; this ...

Analysis

On Mar 24, UTStarcom disclosed OCS prototype development targeted for the second half of 2026; this announcement documented concept prototypes at CIOE, indicating a further development-stage update rather than a disclosed commercial launch.

Key Figures

UOS64 prototype radix: 64x64-radix MOS64 prototype radix: 64x64-radix

UOS64 prototype radix

64x64-radix

Silicon Photonics OCS concept prototype

MOS64 prototype radix

64x64-radix

MEMS-based OCS concept prototype

Historical Context

2 past events · Latest: Mar 24

2 events

1. Mar 24

   AI networking pivot

   24h Move

   -1.5%

   OCS prototype development targeted for the second half of 2026

2. Aug 14

   First-half earnings

   24h Move

   +2.5%

   OCS solution scheduled for showcase at CIOE 2026

24h Move is the share-price change in the day after each event; other market factors may also have contributed.

Key Terms

optical circuit switching, optical-electrical-optical, silicon photonics, photonic integrated circuits, +1 more

5 terms

optical circuit switching technical

"unveiled concept prototypes of its next-generation Optical Circuit Switching"

Optical circuit switching is a network technique that creates a dedicated path for data by routing light signals end-to-end across fiber optics, like reserving a private highway lane so a convoy can travel without stops or slowdowns. For investors it matters because it can dramatically increase raw bandwidth, lower latency and energy use, and simplify data-center and carrier networks — all factors that affect capital spending, operating costs, and the ability to handle surges in traffic.

optical-electrical-optical technical

"eliminating power-intensive Optical-Electrical-Optical (O-E-O) conversions"

A process in fiber-optic networks where an optical light signal is converted into an electrical signal, processed or regenerated, and then turned back into light for further transmission. Think of it like digitizing, cleaning up, and re-broadcasting a radio broadcast: it fixes signal quality and enables switching, monitoring, or protocol changes inside the network. Investors care because O-E-O equipment affects network performance, cost, latency and upgrade flexibility, which influence telecom capital spending and service economics.

silicon photonics technical

"UOS64 (Silicon Photonics Platform)"

Silicon photonics is the technology that uses tiny structures etched into silicon chips to generate, control and detect light for moving data and sensing, essentially putting optical fiber functions onto a computer chip. For investors, it matters because it can dramatically increase data speed and energy efficiency in data centers, telecom networks and advanced sensors, potentially lowering costs and enabling new products much like replacing many metal wires with faster, low-power optical highways.

photonic integrated circuits technical

"chip-scale Photonic Integrated Circuits (PICs)"

Photonic integrated circuits are chips that use tiny optical components to generate, route and detect light instead of moving electrical signals, performing functions similar to electronic computer chips but with photons. They matter to investors because they can enable much faster data transfer and lower power use in telecommunications, data centers and sensing, creating new markets and potential cost advantages — but they also require specialized manufacturing and carry technology and supply risks.

micro-electro-mechanical systems technical

"built on Micro-Electro-Mechanical Systems (MEMS)"

Tiny machines built on silicon that combine mechanical parts, sensors and electronics on a single chip to perform physical tasks like sensing motion, controlling valves, or timing signals. Investors care because these devices enable compact, low-cost products across industries (consumer gadgets, automotive safety, medical devices and industrial sensors), and companies that can design or mass-produce reliable MEMS can scale sales and margins much like a factory that makes a widely used smartphone component.

AI-generated analysis. How Rhea-AI works. Not financial advice.

See more from StockTitan in Google Search and AI answers. Adds StockTitan as a preferred source · opens Google

Add on Google Not now

HANGZHOU, China, Sept. 21, 2026 (GLOBE NEWSWIRE) -- UTStarcom Holdings Corp. (“UT” “UTStarcom” or the “Company”) (NASDAQ: UTSI), a global networking solutions provider, made its inaugural appearance at the 27th China International Optoelectronic Exposition (CIOE 2026), held September 9–11 at the Shenzhen World Exhibition & Convention Center. Exhibiting from a dedicated booth, UTStarcom highlighted its extensive optical networking expertise and unveiled concept prototypes of its next-generation Optical Circuit Switching (OCS) platform designed specifically for AI data center (AI DC) networking infrastructure.

Among a wide variety of optoelectronic applications, CIOE 2026 served as a global stage for leading optoelectronic and optical communications innovators addressing the demanding computing requirements of the AI era. As AI clusters continue to scale, traditional copper interconnects and electrical packet switches (EPS) increasingly encounter power, latency, and thermal constraints. The OCS solutions presented by UTStarcom at the event focus directly on dismantling these barriers by introducing a dynamic, all-optical switching layer into the fabrics of AI data centers.

By eliminating power-intensive Optical-Electrical-Optical (O-E-O) conversions and packet processing at the corresponding AI network layers, UTStarcom’s integrated OCS solution addresses these scaling constraints across three dimensions:

• Enhanced AI Cluster Performance: Operating alongside traditional EPS in hybrid AI data center architectures, OCS provides an automated, dynamically reconfigurable optical layer. By offloading high-volume AI traffic onto direct photonic paths while enabling dynamic topology changes, resource pooling, and optical-layer redundancy, it eliminates buffer-related queuing delays and cuts packet loss and latency during large-scale model training and inference.
• Lower CAPEX and OPEX: Reduces the number of required optical transceivers while lowering power consumption, cooling requirements, and rack footprint.
• Protocol- and Bit-Rate-Agnostic Scalability: Provides protocol- and bit-rate-transparent fiber channels that can accommodate future increases in optical line rates without replacing the physical switching fabric.
  
  

Public Debut of UTStarcom’s OCS Concept Prototypes

During the exposition, UTStarcom unveiled the concept prototypes of its OCS platform alongside its in-house software management ecosystem. The platform remains in active development and comprises the following elements:

• UOS64 (Silicon Photonics Platform): A 64x64-radix OCS concept prototype based on Silicon Photonics (SiPho) technology. Engineered around chip-scale Photonic Integrated Circuits (PICs), it is designed to provide fast optical switching, high-density integration, and scalable optical path automation.
• MOS64 (MEMS Platform): A 64x64-radix OCS concept prototype built on Micro-Electro-Mechanical Systems (MEMS) micro-mirror core technology, designed to deliver low insertion loss, minimal channel crosstalk, and fast optical path switching.
• OCS platform software & SOO Station OCS management platform: UTStarcom’s SONiC-based OCS platform software, paired with the Company’s SDN controller. Through open Southbound and Northbound Interfaces (SBI/NBI), the software layer allows upper-layer orchestrators to automate optical-path provisioning, implement topology-on-demand and streamline optical fabric management.
  
  

Executive Commentary

Dr. Lingrong Lu, Chief Technology Officer of UTStarcom, commented: "Addressing the massive data workloads of next-generation AI clusters requires a fundamental rethink of networking strategies. By bringing both Silicon Photonics and MEMS-based OCS prototypes to CIOE 2026, we demonstrated our vision for a transparent, highly automated optical layer. Exhibiting these concept prototypes allowed us to gather vital technical feedback directly from potential customers and ecosystem partners, ensuring our ongoing R&D roadmap aligns seamlessly with real-world AI deployment requirements."

Mr. Hua Li, Chief Executive Officer of UTStarcom, added: "The rapid expansion of AI infrastructure represents a transformative growth opportunity for optical networking. CIOE 2026 confirmed strong industry interest in flexible, reliable OCS platforms, and the engagement at our booth validated our strategic decision to prioritize AI DC solutions. UTStarcom remains fully committed to continuing investments into product development and working alongside global partners to bring a production-ready OCS portfolio to market."

About UTStarcom Holdings Corp.

UTStarcom provides innovative, reliable, and cost-effective networking solutions worldwide. The Company’s high-performance product portfolio spans high-growth networking domains, including AI data center network infrastructure, telecom mobile backhaul, metro aggregation, and broadband access. UTStarcom is headquartered in Hangzhou, China, and supports operations and customers around the world. UTStarcom was founded in 1991 and listed on the Nasdaq Stock Market since 2000 (symbol: UTSI). For more information about UTStarcom, please visit http://www.utstar.com.

For investor and media inquiries, please contact:

UTStarcom Holdings Corp.
Tel: +86 571 8192 8888
Ms. Shelley Jiang, Investor Relations
Email: utsi-ir@utstar.com /  shelleyjiang@utstar.com

FAQ

AI-generated questions and answers. How Rhea-AI works. Not financial advice.

What are the main components of UTStarcom’s OCS platform unveiled at CIOE 2026?

The OCS platform in development comprises three elements: UOS64, a 64x64-radix Silicon Photonics-based OCS prototype using photonic integrated circuits; MOS64, a 64x64-radix MEMS micro-mirror-based OCS prototype designed for low insertion loss and minimal crosstalk; and the OCS platform software with the SOO Station management platform, which is SONiC-based and works with UTStarcom’s SDN controller to automate optical-path provisioning and optical fabric management.

How is UTStarcom positioning its OCS solution for AI data center networks?

The company positions its OCS solution as an all-optical switching layer that operates alongside traditional electrical packet switches in hybrid AI data center architectures. It is designed to offload high-volume AI traffic onto direct photonic paths, enable dynamic topology changes and resource pooling, reduce packet loss and latency during model training and inference, and lower CAPEX and OPEX by reducing optical transceivers, power consumption, cooling needs, and rack space.

What role does the OCS software and SOO Station management platform play?

The SONiC-based OCS software, together with UTStarcom’s SDN controller and SOO Station management platform, provides open Southbound and Northbound interfaces. This software layer allows upper-layer orchestrators to automate optical-path provisioning, support topology-on-demand, implement optical-layer redundancy, and streamline overall optical fabric management in AI data centers.