With the rapid growth of AI computing infrastructure, 1.6T ultra-high-speed optical modules have entered large-scale commercial deployment, gradually replacing 800G modules and becoming the core hardware for high-speed interconnects in hyperscale data centers.

Compared with previous high-speed optical modules, 1.6T products double the transmission bandwidth and significantly increase throughput. However, these ultra-high data rates also introduce new challenges in signal integrity and stability.

To ensure the long-term reliability of AI computing networks, optical module testing standards have been significantly upgraded. In addition to conventional TX Eye Diagram and TDECQ measurements, Overshoot and Undershoot tests have been introduced as essential verification items. These new tests address transient waveform distortions that were previously difficult to detect and have become mandatory during 1.6T optical module production testing.

I. Why Does 1.6T Require New Dedicated Tests?

For previous-generation 400G and 800G optical modules, signal quality could be adequately evaluated using Eye Diagram masks and TDECQ measurements alone, without dedicated Overshoot or Undershoot testing.

The reason is straightforward: lower-speed electrical signals transition more gradually, making transient waveform abnormalities extremely rare and having little impact on system performance.

The transition to 1.6T transmission fundamentally changes this situation.

According to the IEEE 802.3dj specification, 1.6T optical modules employ 112 Gbaud PAM4 modulation, delivering 224 Gb/s per lane. Eight electrical lanes are aggregated to achieve an overall bandwidth of 1.6 Tb/s, with a Unit Interval (UI) of only 8.92 ps.

Such an extremely narrow timing window and ultra-fast signal transitions mean that even slight hardware imperfections or parameter deviations can cause waveform distortion.

Traditional testing methods mainly evaluate steady-state signal quality and therefore have several limitations:

• Limited measurement scope: Only steady-state signal characteristics at sampling points are measured, making transient waveform distortions invisible.

• Incomplete risk coverage: Overall signal degradation is evaluated, while transient voltage spikes that may damage chips or optical components are overlooked.

• Insufficient stress evaluation: Signal degradation under extreme operating conditions, such as high/low temperatures or heavy workloads, may remain undetected.

• Limited long-term reliability assessment: Existing tests identify immediate bit errors but provide little insight into long-term aging and reliability.

II. Understanding Overshoot and Undershoot

Overshoot and Undershoot are the most common transient waveform abnormalities in 1.6T high-speed transmission. They can be understood as signal “overcorrection” and “undercorrection” during voltage transitions.

Overshoot: During a rising edge, the signal temporarily exceeds its intended high voltage level, producing an excessive positive voltage spike.

Undershoot: During a falling edge, the signal temporarily drops below its intended low voltage level, creating an excessive negative voltage dip.

In practical 1.6T applications, these waveform anomalies are mainly caused by four factors:

• Impedance discontinuities: PCB traces, vias, and via stubs generate high-frequency reflections.

• Connector mismatch: Contact interfaces introduce impedance discontinuities and ringing effects.

• Over-aggressive equalization: Excessive pre-emphasis or equalization sharpens signal edges and produces overshoot.

• High-frequency interference: Power supply noise and crosstalk contribute to transient signal distortion.

III. Core Benefits of the New Tests

Adding Overshoot and Undershoot measurements is not merely expanding the test flow—it represents a fundamental enhancement to the 1.6T signal integrity verification methodology by extending testing from static measurements to comprehensive static + dynamic analysis.

The key benefits include:

• Eliminating testing blind spots by detecting transient waveform distortion that conventional measurements cannot capture.

• Protecting hardware components by preventing excessive voltage spikes from accelerating device aging or damaging optical and electrical ICs.

• Improving transmission stability through more accurate signal levels, resulting in lower BER and packet loss.

• Enhancing product quality by exposing potential failures under stress patterns before mass production.

IV. Industry Standards and Practical Significance

Industry specifications such as IEEE 802.3dj now recognize Overshoot and Undershoot measurements as mandatory verification items for 1.6T optical modules.

The general industry requirement specifies that both Overshoot and Undershoot should remain below 22%. However, leading AI infrastructure providers and hyperscale data centers often enforce stricter internal limits of 10%–15% to maximize link reliability.

These tests complement each other effectively. Eye Diagram and TDECQ evaluate average transmission quality, focusing on inter-symbol interference (ISI), random jitter, and noise. Overshoot and Undershoot measurements evaluate dynamic signal behavior, ensuring long-term hardware reliability.

Overall, the introduction of Overshoot and Undershoot testing is a natural evolution of optical module qualification for ultra-high-speed AI networks.

By comprehensively identifying waveform abnormalities and optimizing hardware design, parameter tuning, and operating conditions, these tests provide a stronger hardware foundation for the high-speed, low-power, and highly reliable operation of AI computing clusters and hyperscale data centers.

0 1

July 7, 2026, Shenzhen, China. – GIGALIGHT today announced that it has been granted two new patents by the China National Intellectual Property Administration (CNIPA).

Patent 1: Optical Path Design for Optical Components Based on a Hybrid DSP and Semi-Linear Direct-Drive Optical Module

Completely different from the conventional design approach for high-speed 800G/1.6T optical modules, this patented technology serves as the core engine and the “heart” of GIGALIGHT’s Hybrid optical module architecture. The innovative optical path design enables the full functionality and complete implementation of the Hybrid optical module.

In general, LRO optical modules share a similar design philosophy with standard FRO optical modules. In contrast, the Hybrid architecture adopts a fundamentally different component layout and signal integrity design compared with FRO.

Like LRO, Hybrid is also a semi-DSP architecture for AI optical interconnects. In an LRO module, the transmitter incorporates a DSP while the receiver is DSP-free. The Hybrid architecture introduces a new framework in which either the transmit or receive path of a single optical module consists of 50% DSP-enabled channels and 50% DSP-free channels. The Hybrid architecture further supports heterogeneous breakout interconnect models for 2×400G and 2×800G optical modules. More importantly, GIGALIGHT views Hybrid as a system-level collaborative compensation architecture that enhances overall link robustness, rather than simply an evolution of LRO technology.

Patent 2: An External OSFP Optical Engine Compatible with NPO/CPO Architectures

ELSFP is currently the industry-standard form factor for external optical engines. This patented external optical engine adopts an OSFP package and introduces an alternative fiber routing method compared with ELSFP. The key advantage is that it enables system architectures with mixed pluggable port configurations, providing greater flexibility for network deployment and system design.

About GIGALIGHT

As an open optical networking explorer, GIGALIGHT integrates the design, manufacturing, and sales of both active and passive optical devices and subsystems. The company’s product portfolio includes III-V optical modules, silicon photonics modules and silicon-based NPO/CPO engines, liquid-cooled optical modules, passive optical components, Active Optical Cables (AOCs), Direct Attach Cables (DACs), coherent optical communication modules, OPEN DCI BOX subsystems based on coherent and O-band DWDM optical modules, and UHD SDI video optical transceivers. GIGALIGHT focuses on applications including AI data centers, 5G transport networks, metropolitan WDM transmission, and ultra-HD broadcast and video, positioning itself as an innovative designer of high-speed optical interconnect hardware solutions.

0 2

June 24, 2026 Shenzhen, China. – The International Conference for High Performance Computing, Networking, Storage and Analysis (ISC High Performance 2026) is taking place from June 23 to 25, 2026, at the Congress Center Hamburg, Germany. ISC High Performance 2026.

GIGALIGHT is exhibiting at the event with its latest 400G/800G AI data center interconnect solutions and immersion liquid cooling solutions. Visitors and global partners are warmly invited to visit Booth G02 for discussions and exchanges.

Featured Solutions :

400G & 800G AI & DC Copper Interconnect Solutions

• 800G OSFP / QSFP-DD PCC / ACC
• 800G OSFP to 2×400G QSFP112 / OSFP-RHS PCC/ACC
• 400G QSFP112 / OSFP-RHS PCC / ACC
• 800G OSFP HYBRID ACC+

400G & 800G AI & DC Optical Interconnect Solutions

• 800G OSFP / QSFP-DD VR8 AOC
• 800G OSFP VR8 to 2×400G QSFP112 / OSFP-RHS VR4 AOC
• 400G QSFP112 / OSFP-RHS VR4 AOC
• 800G QSFP HYBRID VR8 AOC

Immersion Liquid Cooling Solutions

• 800G / 400G / 200G / 100G / 25G Immersion Liquid Cooling Optical Transceivers
• SFP / QSFP / OSFP / QSFP-DD Immersion Liquid Cooling Extenders

Exhibition Highlights

About GIGALIGHT

As an open optical networking explorer, GIGALIGHT integrates the design, manufacturing, and sales of both active and passive optical devices and subsystems. The company’s product portfolio includes III-V optical modules, silicon photonics modules and silicon-based NPO/CPO engines, liquid-cooled optical modules, passive optical components, Active Optical Cables (AOCs), Direct Attach Cables (DACs), coherent optical communication modules, OPEN DCI BOX subsystems based on coherent and O-band DWDM optical modules, and UHD SDI video optical transceivers. GIGALIGHT focuses on applications including AI data centers, 5G transport networks, metropolitan WDM transmission, and ultra-HD broadcast and video, positioning itself as an innovative designer of high-speed optical interconnect hardware solutions.

0 3

June 23, 2026, Shenzhen, China. – GIGALIGHT recently announced the release of its white paper dedicated to Networking Technologies for Tool Intelligence and National Computing Networks. This white paper presents GIGALIGHT’s forward-looking technical positioning and industry outlook, serving as an internal guideline for the company’s differentiated product and technology R&D. Most products and technologies covered in the document have obtained patent authorizations from the National Intellectual Property Administration of China.

The outline of GIGALIGHT’s White Paper on Networking Technologies for Tool Intelligence and National Computing Networks is listed below:

I.Building Low-Latency and Low-Power Data Center Interconnect Networks with NRZ Modulation and Half-DSP Technologies
1.NRZ Modulation: Mature and Reliable for Ultra-Low Latency Performance
2.Half-DSP Modulation: LPO and Hybrid Optical Architectures

II.O-Band Non-Coherent DWDM as a Complement to Coherent Communication for Medium-Reach DCI

III.Three Flexible 1.6T Optical Interconnect Architectures for AI and Data Center Applications
1.Standard 1.6T Pluggable Optical Transceivers
2.External Light Source 1.6T Pluggable Optical Module
3.External Light Source 1.6T (16×100G) NPO Solution Near-Package Optics (NPO)

IV.Ultra-Low-Latency All-Optical Switching at Key Network Nodes Using AWGR-Based Architectures
1.Arrayed Waveguide Grating Router (AWGR) Technology
2.AI Cluster Architecture with AWGR + EDFA
3.AI Cluster Architecture with AWGR + SOA

Full text can be accessed via GIGALIGHT’s official website . Comments and suggestions from industry peers are highly welcome.

About GIGALIGHT

As an open optical networking explorer, GIGALIGHT integrates the design, manufacturing, and sales of both active and passive optical devices and subsystems. The company’s product portfolio includes III-V optical modules, silicon photonics modules and silicon-based NPO/CPO engines, liquid-cooled optical modules, passive optical components, Active Optical Cables (AOCs), Direct Attach Cables (DACs), coherent optical communication modules, OPEN DCI BOX subsystems based on coherent and O-band DWDM optical modules, and UHD SDI video optical transceivers. GIGALIGHT focuses on applications including AI data centers, 5G transport networks, metropolitan WDM transmission, and ultra-HD broadcast and video, positioning itself as an innovative designer of high-speed optical interconnect hardware solutions.

0 4

Computing power is a tool rather than intelligence. Alternatively, we ought to redefine a new type of tool intelligence. Aristotle’s Organon was not composed for everyday utensils. It establishes definitions of categories and logic. Why are categories and logic vital to all human undertakings? Because categories and logic constitute reflection upon thought itself. Adopting a tool-oriented mindset allows us to navigate toward sound scientific and philosophical trajectories.

The national computing power grid demands clear technical insights. Such insights must support upgradability and cross-technology compatibility. To prevent sound computing infrastructure from quickly becoming obsolete amid technological iterations, we must endow such infrastructure with inherent advancement. This advancement belongs to the unified, indivisible category. If we merely follow conventional, non-innovative development paths, the computing power grid will evolve into a consumptive, bloated architecture. Such a framework cannot guide us toward historically rigorous modes of thinking and risks stagnating subsequent development work. Since we define the computing power grid as embodying a degree of tool intelligence, we must operate and unleash this intelligence at minimal cost.

When a person lacks sufficient rest, their mind grows sluggish and briefly stalls or loses vitality. By analogy, the computing power grid resembles the human brain: it must adopt a lightweight daily architecture built for low latency and low-cost innovation, while boasting inherent competitive edges over networking systems deployed in the United States.

I. Building Low-Latency and Low-Power Data Center Interconnect Networks with NRZ Modulation and Half-DSP Technologies

Within AI computing data center networks, NRZ (Non-Return-to-Zero) modulation and “Half-DSP” / linear-drive solutions, including LPO (Linear Pluggable Optics) and Hybrid architectures, have become important technology approaches for achieving both low latency and low power consumption.

These two technology paths are designed for different data rates and transmission distance scenarios. The core objective is to meet the requirements of high-speed interconnects while addressing the high power consumption and latency challenges associated with traditional full-DSP optical modules.

1. NRZ Modulati2. “Half-DSP” Modulation Solutions: LPO and Hybrid Technologiese, Reliable, and Optimized for Ultra-Low Latency

NRZ is a traditional binary modulation technology where each symbol carries one bit of information. In short-distance interconnect applications, its simple physical-layer characteristics make it a preferred solution for latency-sensitive networks.

• Application Scenarios

NRZ is mainly used in interconnect scenarios with single-channel speeds ≤200Gbps.

With continued technology evolution, some high-performance SerDes solutions are exploring NRZ-based optimization approaches for higher-speed applications. However, mainstream high-speed optical interconnects at 400G/800G and beyond have transitioned to PAM4 modulation.

• Key Advantages

Ultra-Low Latency
NRZ does not require complex digital signal processing (DSP) for signal equalization and error correction. The electrical-to-optical conversion process introduces almost no additional processing latency.

Low Power Consumption
With a simpler circuit architecture and without the high power overhead of DSP chips, NRZ-based optical modules typically maintain low power consumption levels, generally within several watts.

High Signal-to-Noise Ratio Tolerance
In optical fiber transmission, channel loss is relatively low. The two-level NRZ signal provides strong resistance to interference and maintains good signal integrity.

• Limitations

NRZ has relatively low spectral efficiency and cannot efficiently support high-speed transmission beyond 50Gbps per wavelength.
Therefore, in mainstream 400G/800G and higher-speed data center and AI computing interconnect architectures, NRZ is gradually being replaced by PAM4 modulation.

• GIGALIGHT’s 200G QSFP-DD S8/PSM8 parallel optical module represents a milestone product in this field.

2. “Half-DSP” Modulation Solutions: LPO and Hybrid Technologies

“Half-DSP” is not a formal industry standard term. It generally refers to reduced-DSP or DSP-free optical architectures, including:

LPO (Linear Pluggable Optics)
Hybrid / LRO (Linear Receive Only)

These solutions aim to maintain sufficient signal integrity while significantly reducing the power consumption and latency associated with traditional full-DSP optical modules.

LPO (Linear Pluggable Optics)

The LPO architecture completely removes the DSP chip from the optical module and retains only the linear driver and TIA (Transimpedance Amplifier).Signal processing and equalization are mainly handled by the advanced equalization capability of the switch-side ASIC.

Hybrid (Half Retimed & Half Linear) / LRO (Linear Receive Only)

Hybrid solutions typically retain partial linear-drive functionality at the transmitter or receiver side, or utilize simplified DSP logic in specific signal paths to achieve a balance between performance and power efficiency.

• Key Characteristics

Balanced Power Efficiency and Performance
Compared with LPO, Hybrid solutions can provide improved signal integrity for longer transmission distances (such as 100m–2km) while still achieving approximately 20%–30% power savings compared with full-DSP solutions.

Latency Optimization
Although some signal processing capability is retained, simplified algorithms or reduced DSP functions enable significant latency reduction.

• Application Scenarios

Hybrid and LRO solutions are suitable for 100m–2km short-to-medium distance interconnect applications, where transmission distance requirements remain important while power efficiency is highly valued.

• Core Technology Comparison

• GIGALIGHT’s Half-DSP / Hybrid optical modules and active optical cable (AOC) product portfolio represent engineering innovations in this field, providing practical solutions for large-scale network deployment with enhanced feasibility compared with traditional LPO/LRO approaches.

Representative products include:

HYBRID Single-Mode Optical Modules

HYBRID Multimode and Single-Mode AOC

HYBRID Copper Cable Solutions

II. O-BAND Non-Coherent DWDM as an Efficient Complement to Coherent Communication for Low-Latency and Cost-Effective Medium-Distance DCI

With the rapid growth of AI computing clusters and distributed data centers, Data Center Interconnect (DCI) networks are facing increasing demands for 400G/800G/1.6T high-speed optical interconnect solutions.

Although traditional coherent communication technologies provide excellent long-distance transmission capabilities, their complex DSP architecture, high power consumption, higher cost, and increased transmission latency have created efficiency and cost challenges for short- and medium-distance DCI applications within 30km.

To address this challenge, GIGALIGHT has introduced an O-BAND (1310nm window) non-coherent DWDM product portfolio.

By leveraging the natural zero-dispersion characteristics of the O-BAND window, this architecture eliminates the need for DCM (Dispersion Compensation Module) and enables an efficient transmission solution optimized for medium-distance data center interconnect applications.

Compared with traditional C-BAND coherent solutions, O-BAND non-coherent DWDM provides the following key advantages:

• Low Latency: By eliminating complex DSP-based dispersion compensation processing, the solution significantly reduces link transmission latency, making it highly suitable for latency-sensitive applications such as AI training and Real-time computing resource scheduling .

• Low Power Consumption: By adopting Silicon Photonics integration technology and PAM4 modulation architecture, the overall power consumption is significantly lower than traditional coherent optical modules.

• Cost Efficiency: The architecture eliminates the need for expensive coherent components, DCM modules, and complex optical line system designs, significantly reducing: CAPEX, Network deployment cost and Operation and maintenance expenses.

• High-Density Deployment: The solution supports high-density optical module form factors including QSFP-DD and OSFP,meeting the requirements of next-generation large-scale data center deployments.

• Open and Flexible Architecture: The solution supports open optical networking architectures and is compatible with white-box switches , open optical line systems and Improving overall network flexibility and deployment efficiency.

From 100G to 800G and further toward 1.6T, GIGALIGHT has established one of the industry’s most complete O-BAND DWDM pluggable product portfolios, covering different generations and capacity requirements for data center interconnect networks.

Product Portfolio:

1. 100G QSFP28 DWDM1

• Single wavelength 100G transmission
• Simplified deployment and operation
• System supports up to 16-channel wavelength transmission

2. 200G/400G QSFP-DD DWDM4

• 4-wavelength aggregation
• 50G/100G per wavelength
• System supports up to 16-channel wavelength transmission

3. 800G OSFP/QSFP-DD 2×PSM DWDM4 (O-BAND)

• Single wavelength 100G PAM4
• System supports up to 16-channel wavelength transmission

4. 1.6T OSFP224-EL 2×PSM DWDM4 (O-BAND)

• Single wavelength 200G PAM4
• External light source architecture
• System supports up to 16-channel wavelength transmission

O-BAND DCI Interconnect Application Scenario

Transmission Distance: 2km–30km

III. Three Flexible Solutions for 1.6T AI & Data Center Interconnect

With the explosive growth of AI computing requirements, data center interconnect bandwidth is rapidly evolving from 400G/800G toward 1.6T.

To meet deployment requirements across different application scenarios, the industry is developing three major technology approaches for 1.6T optical interconnects. Each solution provides different advantages and is optimized for different requirements in terms of cost structure ,power budget and deployment scalability.

1. Standard 1.6T Pluggable Optical Transceiver

Standard 1.6T OSFP224 Optical Module

The 1.6T OSFP224 optical module is a next-generation high-speed optical transceiver featuring:

• Total bandwidth of 1.6Tbps
• OSFP224 form factor
• Designed to support AI clusters and high-performance computing bandwidth requirements

It represents one of the mainstream solutions for next-generation AI networking deployments.

Core Parameters and Definition

Data Rate and Form Factor

Total Data Rate : 1.6Tbps total bandwidth is achieved through 8 × 200G PAM4 electrical channels. After protocol overhead, the effective payload reaches approximately 1.6Tbps.

OSFP224 Form Factor：OSFP224 is an evolution of the traditional OSFP form factor. It features:

• 224 electrical contacts
• Support for 16 electrical lanes
• Higher port density
• Enhanced thermal capability

It is designed for high-density AI switch deployments.

Standard 1.6T Pluggable Module Solutions

2.External Light Source 1.6T Pluggable Optical Module (External Light Source Pluggable, ELS-Pluggable)

This architecture separates the laser source from the 1.6T OSFP224 optical module. A centralized:

• Rack-level external light source
• Board-level external light source
• Co-packaged light source

provides optical power to multiple pluggable modules through fiber coupling.

The optical module itself only retains modulators and photodetectors and required electrical processing components. This architecture significantly reduces thermal load at the individual module level.

As one of the pioneers proposing the External Light Source 1.6T Pluggable Optical Module architecture, GIGALIGHT provides the following solutions:

3. External Light Source 1.6T (16×100G) NPO Solution Near-Package Optics (NPO)

The NPO architecture represents a transition technology between pluggable optics and Co-Packaged Optics (CPO). In the NPO architecture, the optical engine is no longer integrated inside the switch ASIC package.

Instead, it is installed immediately outside the package：Optical engine distance from ASIC: <50mm

Combined with an external light source, NPO enables ultra-short-distance high-speed interconnects through Silicon Photonics or Polymer waveguide technologies.

The 1.6T NPO solution is typically deeply integrated with Silicon Photonics platforms to achieve extremely low electrical-to-optical transmission loss.

It should be noted that practical NPO optical engine architectures generally need to be based on:

16×100G physical architecture
32×100G physical architecture

This represents the key differentiation point compared with 1.6T pluggable optical modules.

Although the industry is exploring 3.2T NPO and 6.4T NPO

Architectures based on 224G SerDes technology, related standards have not yet been finalized.

Comparison of Three 1.6T Solutions

Deployment Roadmap

The three solutions are not mutually exclusive alternatives.

Instead, they represent a layered deployment strategy targeting different technology maturity stages and application scenarios.

Near Term (2024–2026): Standard pluggable optical modules will remain the mainstream solution, supporting smooth upgrades for most data center deployments.

Mid Term (2027+): External light source pluggable architectures are expected to achieve rapid adoption in large-scale AI clusters by balancing deployment flexibility and power optimization

Long Term (2027+): With the maturity of Silicon Photonics technology and standardization progress, NPO solutions are expected to become a major architecture for AI supercomputing interconnects, providing the foundation for future 3.2T and 6.4T network evolution.

IV. Enabling Ultra-Low Latency All-Optical Switching at Critical Network Nodes with AWGR + EDFA / AWGR + SOA Architectures

1. AWGR (Arrayed Waveguide Grating Router)

AWGR is a passive wavelength routing engine that utilizes its cyclic wavelength routing characteristics to deterministically map different optical wavelengths within a WDM signal to corresponding output ports.

Unlike traditional electrical switching architectures, AWGR does not require lookup tables, packet scheduling and optical buffering.

Therefore, it completely eliminates the traditional optical-electrical-optical (O-E-O) conversion process in electrical switching systems.

The optical signal remains entirely in the optical domain throughout the transmission process.

A single-stage optical switching operation introduces only nanosecond-level latency, which is:

• 2–3 orders of magnitude lower than traditional OEO architectures
• Highly deterministic with extremely low latency jitter

This capability makes AWGR-based optical switching highly suitable for:

• AI computing clusters
• Low-latency computing networks
• High-performance computing scheduling applications

2. AI Computing Cluster AWGR + EDFA Optical Switching Topology

1) Architecture Overview

This architecture is mainly designed for data center inter-rack optical switching and cross-campus long-distance all-optical switching hubs.The topology adopts a three-level flattened architecture:

Compute Access Layer → Core Wavelength Switching Layer → Optical Amplification Output Layer

Access Layer

• Each AI computing node / GPU server outputs optical signals through:
• 100G PSM DWDM4 (C-BAND) optical modules
• Each optical port directly carries:
• Four fixed C-band wavelengths
• The optical signals are then aggregated through a:
• DWDM multiplexer (MUX)

Core Switching Layer

• After wavelength multiplexing, signals are transmitted into an:
• N×N AWGR wavelength router
• The AWGR performs optical routing switching based on wavelength assignment.
• The switching process is completed entirely in the optical domain.

Optical Compensation Output Layer

• After AWGR switching, the system experiences approximately:
• 8dB inherent insertion loss
• An EDFA (Erbium-Doped Fiber Amplifier) is used to compensate for this optical loss.
• The amplified signal is then transmitted to:
• Next-stage switching nodes
• Target computing clusters
• Supporting transmission distances at the:
• 10km level

2) Topology Characteristics

Key advantages include:

• No multi-stage O-E-O conversion
• Core switching latency below 1μs
• Simplified network architecture
• Suitable for large-scale computing hubs with relatively stable network topology

Technical Parameter Comparison

3. AI Computing Cluster AWGR + SOA Optical Switching Topology

This architecture is mainly designed for:

• In-rack optical switching
• Same data center high-density optical switching hubs
• Ultra-low latency AI cluster interconnects

The topology adopts a compact architecture combining:Active Scheduling + All-Optical Switching + Integrated Optical Amplification

Wavelength Scheduling Layer

Each computing node connects to:

• O-BAND DWDM optical components
• O-BAND multi-wavelength tunable optical engines

The scheduling controller dynamically adjusts output wavelengths according to GPU communication requirements, including Gradient synchronization and Parameter broadcasting.

GIGALIGHT provides O-BAND DWDM parallel optical modules with different baud rates to support medium-distance AI computing interconnect applications.

Core Optical Switching Layer

The dynamically scheduled wavelength signals are transmitted into:

AWGR wavelength routers

The AWGR performs:

• Wavelength-based optical routing
• Nanosecond-level optical path switching

Integrated Optical Amplification Layer

The architecture uses:

SOA (Semiconductor Optical Amplifier) for amplification of weak optical signals after switching.

Compared with EDFA, SOA provides:

• Smaller form factor
• Better integration capability
• Suitability for high-density rack deployment

SOA can also assist with:

• Wavelength conversion
• Enhanced scheduling flexibility

Topology Characteristics

The complete switching process latency can be reduced to:

• Below 100 nanoseconds
• The power consumption of a single optical switching chip is approximately:
• 1/10 of a comparable electrical switching ASIC

This architecture is highly suitable for:

• Large-scale AI training clusters
• Thousands of GPU interconnect environments
• Low-latency east-west GPU communication

The network architecture diagram

Key Node Selection Recommendations

Conclusion

The core essence of a computing power network resides in its capability to invoke tools. A pure computing power network, an expensive piece of infrastructure, would be utterly meaningless without a society-wide foundation of lifelong learning.

From any perspective of commercial development, blind dependence on large computing models will clash with knowledge undertakings constructed through language. Commercial organizations may only conduct innovative thinking based on computing power networks; they must not regard computing power itself as a competitive advantage.

Computing power networks should function as networks that boost the intellectual capacity of all citizens. Otherwise, trillions of capital invested will fail to deliver any positive outcomes.

Just as traditional telecommunication pipelines were built to connect information, computing power networks exist to connect humans with instrumental intelligence.

Only when we precisely define the core value and positioning of computing power networks can this network become a digital infrastructure worthy of massive national investment.

0 5

June 22, 2026, Shenzhen, China. – GIGALIGHT today officially announced the relocation of its Technology Center and Manufacturing Facility to a brand-new office location.

Dear Customers, Suppliers, and Business Partners,

Greetings!

We sincerely appreciate your long-term trust and strong support for GIGALIGHT. As our business continues to expand, the original office and production space can no longer meet our development needs. Effective June 22, 2026, our Technology Center and Manufacturing Facility have officially relocated to the following new address:

Floors 10–11, Building A, Longbang Kexing Science Park, Guangming District, Shenzhen, Guangdong Province

We look forward to meeting and communicating with you in person at our new office location.

Tel: +86-755-26734300

Email: [sales@gigalight.com]

About GIGALIGHT

As an open optical networking explorer, GIGALIGHT integrates the design, manufacturing, and sales of both active and passive optical devices and subsystems. The company’s product portfolio includes III-V optical modules, silicon photonics modules and silicon-based NPO/CPO engines, liquid-cooled optical modules, passive optical components, Active Optical Cables (AOCs), Direct Attach Cables (DACs), coherent optical communication modules, OPEN DCI BOX subsystems based on coherent and O-band DWDM optical modules, and UHD SDI video optical transceivers. GIGALIGHT focuses on applications including AI data centers, 5G transport networks, metropolitan WDM transmission, and ultra-HD broadcast and video, positioning itself as an innovative designer of high-speed optical interconnect hardware solutions.

0 6

As AI large model computing infrastructure accelerates across the industry, with 800G achieving large-scale deployment and 1.6T entering mass commercialization, the power consumption of high-speed optical modules continues to surpass the physical limits of air-cooling. Liquid-cooled optical modules have been upgraded from customized optional solutions to standard products for high-end AI data centers, becoming a critical necessity in computing network construction. Leveraging the superior heat transfer coefficient of liquid, these modules address key challenges such as frequency throttling and rising bit error rates caused by extreme heat in ultra-high-bandwidth devices.

I. Definition and Mainstream Technical Solutions of Liquid-Cooled Optical Modules

Liquid-cooled optical modules are specialized optical interconnect devices developed for high-density AI computing clusters. They use coolant to directly dissipate heat from core components such as DSPs and EML lasers, achieving heat transfer efficiency more than three times that of conventional air cooling, enabling reliable continuous operation of high-power devices. Based on the overall thermal management architecture of data centers, the industry has formed three mainstream technical approaches.

Approach 1: Cold Plate Contact Liquid Cooling

This is also the mainstream solution for 800G/1.6T pluggable modules. A fixed copper cold plate is integrated into the switch cage, and after the optical module is inserted, its housing makes tight contact with the cold plate. Internal circulating water cooling removes the heat. This approach has a low retrofit difficulty, is compatible with existing OSFP and QSFP-DD form factors, and suits the upgrade needs of existing data centers transitioning from air cooling to liquid cooling.

Approach 2: Module-Embedded Microchannel Liquid Cooling

Microchannels are pre-embedded inside the optical module housing, paired with quick-connect fluid couplings, allowing coolant to directly contact the heat-generating chips. The thermal dissipation limit per module exceeds 80W, making it compatible with XPO ultra-high-density optical modules. However, the sealing and leak-prevention manufacturing requirements are relatively demanding.

Approach 3: Full Immersion Liquid Cooling

The entire optical module is submerged in fluorinated liquid, enabling uniform heat dissipation across all surfaces. It is suited for ultra-high-power single-rack AI clusters exceeding 60 kW. Small-scale pilot deployments have been conducted domestically, but the approach requires full rack retrofitting, resulting in high upfront infrastructure costs.

Compared to traditional air-cooled optical modules, liquid-cooled products significantly reduce operating temperatures, greatly lowering component aging and link bit error rates caused by excessive heat. They also help data centers achieve a PUE below 1.15, aligning with domestic energy consumption standards for new computing centers. In terms of product specifications, current liquid cooling retrofits prioritize 1.6T and above products, where heat flux density exceeds 100 W/cm² and air cooling is insufficient. The optimized LPO architecture brings 800G power consumption below 10W, allowing air cooling to remain viable in some mid-to-low-end scenarios, with liquid cooling reserved for high-end training clusters.

II. Current Industry Deployment Status

Global AI computing infrastructure investment continues to intensify, with overseas cloud providers and domestic AI data centers maintaining elevated capital expenditures. The explosive demand for high-speed, high-power optical modules is driving liquid cooling solutions from pilot deployments to large-scale commercial adoption.

On the supply side, a domestic-manufacturer-led industry structure has taken shape, with leading enterprises outpacing global peers in technology iteration and mass production. Innolight holds the top position in the global optical module industry, with over 40% global market share in 800G and 50%–70% in 1.6T. Its liquid-cooled 1.6T optical modules have achieved stable volume delivery, with mature cold plate integration manufacturing processes and deep compatibility with overseas high-end computing cluster requirements.

The industry standards ecosystem is also maturing, effectively resolving early challenges of non-standardization and poor compatibility. China’s ODCC has officially released the White Paper on Key Liquid Cooling Technologies for 800G/1.6T Optical Modules, unifying cold plate dimensions, contact methods, fluid interfaces, and sealing specifications for mainstream pluggable optical modules, providing a standards basis for retrofitting existing data centers and scaling up new ones. Meanwhile, leading industry players are jointly advancing frontier standards such as XPO ultra-high-density packaging and next-generation liquid cooling interfaces to accommodate 3.2T and above ultra-high-speed liquid-cooled optical modules, driving the industry’s upgrade from volume production toward standardization, normalization, and high reliability.

III. Existing Industry Development Challenges

Despite the positive industry momentum, the widespread adoption of liquid-cooled optical modules remains constrained by several bottlenecks.

Sealing and leakage reliability: The plug-in connectors for embedded microchannel liquid cooling modules require high precision, and the risk of leakage under long-term thermal cycling is difficult to fully eliminate. Immersion solutions also require all module components to be modified for resistance to fluorinated liquid immersion, increasing material costs by more than 20%.

High supply chain costs: Domestic production rates for thermal components such as copper cold plates, quick-connect fluid couplings, and specialized sealants remain insufficient, with high-end accessories dependent on imports. Liquid-cooled optical modules carry a 25%–40% price premium over air-cooled equivalents at the same speed, limiting adoption rates among small and mid-sized computing centers.

Packaging compatibility conflicts: Traditional QSFP-DD and OSFP cage spaces are compact, leaving insufficient structural room after cold plate installation. The high cost of large-scale retrofitting of existing switches constrains the upgrade pace of legacy data centers.

IV. Core Future Industry Development Trends

Continuous product speed upgrades, with liquid cooling becoming standard for 3.2T: As single-channel 400G electro-optical devices are developed and commercialized, full-scale R&D of 3.2T optical modules has launched. Product power consumption will exceed 60W, at which point air cooling will be entirely replaced by liquid cooling.

Deep coupling of silicon photonics, CPO, and liquid cooling: After co-packaged optics (CPO) and silicon photonics technologies are integrated, local heat flux density multiplies, necessitating embedded microchannel liquid cooling. Within the next 2–3 years, liquid cooling designs are expected to be completed concurrently during the optical engine packaging stage, achieving native thermal optimization.

Accelerating domestic substitution with declining costs: Domestic enterprises in precision copper processing and sealing components are accelerating technology breakthroughs, gradually increasing the domestic production rate of cold plates, fluid couplings, and other components. By 2027, the price gap between liquid-cooled and air-cooled products is expected to narrow to within 10%, and the domestic liquid-cooled optical module market is projected to exceed 10 billion RMB. Additionally, low-power technologies such as LPO will give rise to lightweight semi-liquid-cooled solutions suited for mid-to-low-end data centers.

Gradual scaling of immersion liquid cooling: In the short term, cold plate contact cooling remains dominant. Over the medium to long term, full immersion liquid cooling will accelerate penetration owing to its superior thermal efficiency, shifting data centers from single-point liquid cooling to full-link liquid cooling. Driven by green computing policies, the liquid-cooled optical module industry is expected to enter a high-growth cycle lasting more than five years, with a compound annual growth rate exceeding 50%.

Conclusion

AI computing demand is reshaping the optical communications industry, and the transition of liquid-cooled optical modules from customized to standardized products has become a clear trend. In the short term, 800G/1.6T liquid-cooled products will continue to scale in volume. Over the medium to long term, silicon photonics, CPO, and advanced packaging technologies will drive liquid cooling to become standard across all 3.2T and higher speed products. As domestic substitution reduces costs, liquid-cooled optical modules are poised to become the universal form factor for data communications optical modules, with deep benefits from global computing infrastructure buildout.

0 7

May 26, 2026, Shenzhen, China. – It has recently been learned from the intellectual property authority that GIGALIGHT has been granted a utility model patent for a silicon photonics-based parallel O-BAND DWDM optical module, which was filed in June 2025. The corresponding invention patent application is currently under substantive examination.

O-BAND (defined by the ITU wavelength range of 1260 nm to 1360 nm) parallel DWDM is designed to address the challenges of system cost, power consumption, and latency in high-baud-rate, short- to medium-reach transmission applications.

Before GIGALIGHT introduced its O-BAND non-coherent DWDM transmission architecture, the mainstream approach for high-baud-rate DWDM transmission in the industry was coherent transmission. At the same time, the trend of lightweight coherent technology extending into DCI applications has become increasingly evident. However, the inherent complexity of coherent technology brings challenges related to cost efficiency and latency in DCI interconnection scenarios, making it difficult to achieve an ideal cost-performance balance while also introducing non-negligible system latency.

The O-BAND non-coherent DWDM optical module developed by GIGALIGHT not only provides cost advantages for short- to medium-reach DCI interconnections (although hyperscale data centers and AI infrastructure operators are not always solely focused on cost), but more importantly delivers significant system latency benefits. The greatest value of this solution lies in achieving the industry’s lowest system transmission latency, which is critical for latency-sensitive applications such as AI training and real-time computing.

O-BAND Non-Coherent DWDM: The Optimal Solution for High-Bandwidth Short- to Medium-Reach Transmission.

About GIGALIGHT

As an open optical networking explorer, GIGALIGHT integrates the design, manufacturing, and sales of both active and passive optical devices and subsystems. The company’s product portfolio includes III-V optical modules, silicon photonics modules and silicon-based NPO/CPO engines, liquid-cooled optical modules, passive optical components, Active Optical Cables (AOCs), Direct Attach Cables (DACs), coherent optical communication modules, OPEN DCI BOX subsystems based on coherent and O-band DWDM optical modules, and UHD SDI video optical transceivers. GIGALIGHT focuses on applications including AI data centers, 5G transport networks, metropolitan WDM transmission, and ultra-HD broadcast and video, positioning itself as an innovative designer of high-speed optical interconnect hardware solutions.

0 8

I. Background: Layered Requirements of Computing Networks

With the deployment of large-scale AI training and inference clusters, computing interconnect networks are developing a distinct layered architecture:

•  Scale Across Network (short-to-medium range): Interconnects computing nodes between racks, cabinets, and within machine rooms, typically covering distances from hundreds of meters to a few kilometers. Latency-sensitive and extremely cost-sensitive—this is the core of today’s GPU/NPU cluster networking.

•  Computing Backbone Network (long distance): Cross-datacenter, cross-city computing dispatch over tens to thousands of kilometers, requiring spectral efficiency and line amplification. Coherent communication is the indispensable technology here.

These two types of networks have fundamentally different requirements for optical modules. Long-distance computing networks must use coherent communication, whereas Scale Across short-to-medium range networks have absolutely no need for coherent solutions—the additional cost, power consumption, and latency introduced by coherent systems are purely burdensome in this scenario.

II. Why Does Scale Across Choose O-Band Incoherent Solutions?

O-Band (1260–1360 nm) offers three key advantages in short-to-medium range transmission:

•  Zero-Dispersion Window: O-Band operates near the zero-dispersion point of optical fiber—no dispersion compensation is needed, signal integrity is excellent, and it accommodates high-order modulation formats like PAM4 without DSP dispersion equalization overhead.

•  Low Latency: The incoherent architecture eliminates the processing delay of coherent DSP, achieving extremely low end-to-end latency—critical for collective communication operations like All-Reduce in AI training.

•  Most Economical System Cost: No local oscillator laser, 90° optical hybrid, or complex DSP required; silicon photonics solutions are sufficient, and BOM cost is substantially lower than coherent solutions.

In Scale Across scenarios, O-Band incoherent solutions provide the optimal cost-latency-power triangle balance.

III. Current Product Coverage: Up to 800G Without Obstacles

The current incoherent solution based on single-wavelength 100G O-Band can fully cover up to 800G speeds. In the future, the 800G solution can be built on the proven single-channel 200G PAM4, with low technical risk and high system completeness.

IV. 1.6T Scale Across: Evaluation of Three Technical Pathways

When evolving toward 1.6T, there are three viable pathways, each with different technology maturity and engineering constraints:

Path 1: 1.6T PSM DWDM4 (Single-Wave 400G PAM4, Silicon Photonics Modulation)

This is the most aggressive single-module solution. The DSP for single-wave 400G PAM4 has not yet entered mass production, but 1.6T DR4 samples have already appeared in the industry, indicating that 400G PAM4 DSP chip samples exist. The transition from samples to mass production will take time; this is expected to be a viable mid-term direction.

Path 2: 1.6T 2×PSM DWDM4 (Single-Wave 200G PAM4, Silicon Photonics Modulation)

Based on proven 200G PAM4 silicon photonics modulators, achieving 1.6T by stacking two 800G sub-modules. Technically feasible, but power consumption is the core bottleneck: if total power exceeds 30W, switch ports cannot directly power the module, requiring additional power supply design that breaks the plug-and-play nature of pluggable modules. This solution is only suitable for specific subsystem deployment under power constraints and is unlikely to become the mainstream Scale Across solution.

Path 3: 1.6T NPO 2×(PSM DWDM4) (Single-Wave 200G PAM4) ✓ Recommended

The NPO architecture places the optical engine near or integrated into the switch chip package, fundamentally resolving the dual challenges of power consumption and signal integrity. This is the most viable engineering path for achieving 1.6T Scale Across at the current stage.

•  Extremely short electrical signal routing, significantly reducing SerDes power consumption

•  No longer constrained by the power budget of standard pluggable modules

•  Based on proven 2×PSM DWDM4 architecture with a clear technology path

•  Can fully achieve 1.6T O-Band incoherent Scale Across

V. Technology Pathway Comparison Summary

VI. Conclusion

The fundamental requirements of Scale Across computing networks are: low latency, low cost, high density, and easy operation and maintenance. In short-to-medium range computing interconnect scenarios, the system complexity introduced by coherent communication is completely unnecessary, and O-Band incoherent solutions naturally align with these requirements.

•  O-Band incoherent 400G and 800G PSM DWDM4 solutions are technically mature and ready for large-scale deployment.

•  O-Band incoherent 1.6T DWDM can be achieved through design innovation.

0 9

May 11, 2026, Shenzhen, China. – With the rapid acceleration of AI clusters, data center switching architectures, and 200G/400G network deployments, demand for higher port density and more flexible network breakout solutions continues to grow. Dual-density SFP56-DD optical modules are becoming an important option for next-generation high-performance switching and interconnect architectures.

GIGALIGHT introduces the dual-density 100G SFP56-DD SR2 optical module based on 2×50G PAM4 modulation technology, delivering double-density capability and outstanding port density to meet the requirements of next-generation high-performance networks. The SFP56-DD (Double Density) form factor maintains the compact size of the traditional SFP+ module (approximately 14 × 56 mm) while enabling 100 Gbps high-speed transmission through two 53.125 Gbps electrical lanes. Compared with QSFP28 (18.35 × 72.4 mm), the SFP56-DD significantly improves PCB space utilization and reduces board space consumption by approximately 35%. This allows more ports to be deployed within the same system footprint, enabling ultra-high-density network deployments and improved system scalability.

100G SFP56-DD SR2 Product Features

• Form factor: SFP56-DD
• Power consumption: <3.5W, typical 3W
• Optical interface: 2×50G PAM4 VCSEL
• Electrical interface: 2×50G PAM4
• Optical connector: MPO-12/UPC
• Transmission distance: 70m on OM3 MMF and 100m on OM4 & OM5 MMF
• DSP inside
• CMIS V4.0 compliant

Flexible 200G Port Breakout Application Scenarios

GIGALIGHT will continue to focus on innovation in high-speed optical interconnect technologies, providing global customers with higher-density, lower-power, and more flexible optical communication solutions, driving the continuous evolution of AI and data center networks.

About GIGALIGHT

As an open optical networking explorer, GIGALIGHT integrates the design, manufacturing, and sales of both active and passive optical devices and subsystems. The company’s product portfolio includes III-V optical modules, silicon photonics modules and silicon-based NPO/CPO engines, liquid-cooled optical modules, passive optical components, Active Optical Cables (AOCs), Direct Attach Cables (DACs), coherent optical communication modules, OPEN DCI BOX subsystems based on coherent and O-band DWDM optical modules, and UHD SDI video optical transceivers. GIGALIGHT focuses on applications including AI data centers, 5G transport networks, metropolitan WDM transmission, and ultra-HD broadcast and video, positioning itself as an innovative designer of high-speed optical interconnect hardware solutions.

10