The Optical Fiber Communications Conference and Exhibition (OFC) is celebrating 50 years of optical networking and communications this year. The milestone highlights decades of innovation from companies such as Acacia that have delivered optical networking technology enabling network operators to continually expand their bandwidth, deliver faster networks, and offer more services.

One of those innovations that we believe will be a dominant topic at OFC this year is coherent optics. Coherent pluggable modules have been critical for enabling network operators to keep up with bandwidth demands over the last decade, and according to Cignal AI, they will account for most of the future bandwidth growth driven by applications such as generative AI and machine learning. Acacia pioneered the coherent pluggable more than 10 years ago with the introduction of the industry’s first 100G CFP module in 2014 and has been a market leader ever since having shipped half a million 400G ports based on the Greylock DSP, including more than 25,000 Bright 400ZR+ ports.

Acacia continues to innovate and will illustrate its industry-leading optical networking advancements through the following demonstrations and speaking sessions:

Demonstrations:

• 200G per lane optical engine products – Demonstrating the recently announced 200G/lane silicon photonics optical engine with transmitter and receiver designs supporting DR4, DR8 and 2xFR4 configurations.
• Delphi DSP-based Coherent Pluggable Portfolio including new L-Band 400G ULH – Showcasing Acacia’s broad 400G module portfolio powered by Acacia’s 9th generation DSP Delphi including 800ZR, 800G ZR+ with interoperable Probabilistic Constellation Shaping (PCS), and 400G Ultra Long Haul. 400G modules in QSFP-DD form factor and 800G modules in both OSFP and QSFP-DD form factors are available with C and L-Band versions.
• OIF multi-vendor interoperability (Booth #5745) – Showing interoperability of 800ZR, 400ZR, OpenZR+ and Multi-Span Optics enhancing multi-vendor interoperability in high-capacity, long-distance optical networks and Common Management Interface Specification (CMIS) standardizing the management of optical and electrical devices to simplify scalability.
• OpenROADM compliant 800G ZR+ coherent modules – Featuring a link between the OIF booth (#5745) and OpenROADM booth (#5128) demonstrating the high-performance interop PCS interface which expands the market for 800G pluggables beyond metro DCI into regional and even long-haul networks. Acacia’s award-winning 800G ZR+ modules were the first in the industry to support the OpenROADM specification.
• Ethernet Alliance (Booth #5173) – Showing interoperability of Acacia 800G ZR interconnects.

Speaking Sessions:

Monday, March 31

• Optica Executive Forum: Status of Photonic-enabled Modules and Interconnects
  — Speaker: Tom Williams, VP of Marketing
  — Time: 10:35-11:50 am
  — Location: Marriott Marquis
• Multi-rate PAM4 Silicon Photonic Based Receiver Assembly with -10dBm iFEC Sensitivity at 226Gb/s/λ and 125mW/Ch CMOS TIA
  — Speaker: Mahdi Parvizi, Principal Engineer
  — Time: 10:30 am to 12:30 pm
  — Location: Rooms 211-212 (Level 2)

Tuesday, April 1

• Market Watch: Optical Modules, Transceivers and Applications
  — Speaker: Tom Williams, VP of Marketing
  — Time: 2:15 – 3:45 pm
  — Location: Theater I

Wednesday, April 2

• Network Operator Summit: Interoperation of Optical Pluggable Transceivers and IP/Optical Integration
  — Moderator: Anuj Malik, Director, Product Management and Strategy
  — Featured speakers from Cignal AI, Arelion, Cox Communications, Deutsche Telekom Technik GmbH, Telefonica
  — Time: 10:45 – 12:15 pm
  — Location: Theater I
• Coherent Optics Unleashed: 400ZR Success to 800ZR/LR Advancements and 1600ZR/ZR+ Kick-Off
  — Speaker: Doug Cattarusa, Technical Marketing Engineer
  — Time: 4:00 – 5:00 pm
  — Location: Theater I

Thursday, April 3

• Market Watch: Optical Network Evolution for AI/ML, Architectures and Drivers
  — Speaker: Tom Williams, VP of Marketing
  — Time: 10:15 – 11:45 am
  — Location: Theater I
• Demonstration of 400G Optical Interoperability in IP Over WDM Network Scenarios Over Single Span WDM System with and Without Optical Amplifiers
  — Speaker: Anuj Malik, Director, Product Management & Strategy
  — Time: 10:30 am
• Coherent Moving to Client Optics
  — Moderator: Tom Williams, VP of Marketing
  — Time: 2:00 – 3:00 pm
  — Location: Theater II

Come See us!
If you are attending OFC 2025 and would like to discuss any of these topics, contact us to request a meeting.

Driven by Demands for Higher-Bandwidth, Lower Power and Smaller Footprint
Acacia’s client optics portfolio is driven by key customer requirements for better power efficiency, higher performance, and smaller footprint from a proven high-volume supplier. Delivering these capabilities is a critical enabler for module vendors to design cutting-edge pluggable modules that can handle the compute intensive workloads generated by AI, cloud services, and video streaming. With more than a million 100G per lane optical engines shipped in the last 12 months, Acacia is already an established supplier for client optics components based on silicon photonics.

Introducing the 3nm 1.6T Kibo PAM4 DSP
At the heart of Acacia’s 200G per lane client optics portfolio is the 3nm 1.6T Kibo PAM4 DSP designed to power the optical interconnects inside the world’s cloud and AI data centers. It is expected to sample later in 2025.

Key features include:

• 3nm CMOS node for market leading power efficiency enabling more than 20% lower power compared to existing 1.6T module implementations
• Industry standard compliance enhanced by Acacia’s algorithms
• Transmit Retimed Optics (TRO) configurations with power-efficient support for diagnostic and loopback troubleshooting capabilities
• Support for gearbox and retimer applications
• Designed to support 1.6T DR8 and 2xFR4, as well as 800G DR4, DR8, FR4 and 2xFR4 modules in OSFP/QSFP-DD form factors

200G per Lane Optical Engine Family
Complementing the Kibo PAM4 DSP is a family of Optical Engine products designed to support 200G per lane that leverage Acacia’s silicon photonics expertise. Acacia’s Optical Engine products will be demonstrated at OFC 2025.

The family of Optical Engine products address a variety of applications with the following configurations:

• Separate transmit and receive components
• DR4, DR8 and 2xFR4 use cases
• Support for 100G/lane and 200G/lane
• Flexible driver configuration support
• Transimpedance Amplifier (TIA) integration in receiver circuit (RX)

“We expect the market for IC chipsets for optical communications to grow from 2025 through 2030 at a CAGR of 17% as hyperscalers look to meet the burgeoning demands that AI is placing on the entire infrastructure,” said Vladimir Kozlov, Founder and CEO at LightCounting. “Having a new supplier in this space, that also has the credibility that Acacia has amassed over more than a decade, will be key for continuing innovation.”

Leverages Acacia’s Proven Volume Manufacturing Capabilities
As an established leader in coherent DSP ASIC design-to-volume deployment, Acacia is a trusted supplier in the telecommunications industry. Acacia’s expanded client optics portfolio is also backed by Cisco’s full commitment to support the ongoing development, supply and customer service of Acacia’s existing and future products.

“Client optics has always represented an exciting opportunity to leverage our team’s expertise in DSP and silicon photonics,” said Benny Mikkelsen, SVP and General Manager, Acacia. “With AI driving tremendous demand for client optics and increased adoption of silicon photonics at 200G/lane, this is an ideal time for us to invest in scaling this part of our business.”

Come See us At OFC 2025
If you are attending OFC 2025 and want to discuss our new client optics offerings, we’d welcome the opportunity to meet with you. Click here to set up a meeting.

Figure 1. Acacia’s Wall of Patents.

A recent Gazettabyte article details how our co-founders carried out this mission, introducing improvements in performance and power consumption to legacy coherent technology, initially used for the most challenging optical transmission links. Looking through our history, it’s clear to see the trailblazing path Acacia helped to define with coherent optical modules, especially pluggable modules. Acacia has always been laser-focused on providing customers with continuous improvements in speed, capacity, performance, size, and power consumption to meet the ever-growing demand for bandwidth. This has been done by focusing on (1) silicon-based optical and electrical designs, (2) advanced digital signal processing (DSP), (3) high-speed RF expertise, and (4) volume manufacturable high-density packaging, with all these disciplines “under one roof.” As a result, Acacia is now a leading supplier of coherent modules with multiple product families to support speeds from 100G to 1.2T, including a broad portfolio of 400G and 800G pluggable products, to address the needs of network operators worldwide.

Pioneer of Coherent Pluggables
It is interesting to note that this year marks the 30th anniversary of a significant event in optical transceiver history. It was in 1995 when the first standards-based pluggable optical transceiver called the Gigabit Interface Convertor (GBIC) was first defined, with supported speeds up to ~1Gbps. It was ~20 years from this achievement that Acacia introduced the first optical coherent pluggable transceiver to support 100Gbps in the multi-source agreement (MSA) CFP form factor.

Figure 2. Industry standardized pluggable modules: From the early days of 1G GBICs to today’s coherent pluggables capable of 3000km+ 400G and 1000km+ 800G.

And today, Acacia’s latest generation of coherent pluggable modules powered by the Delphi DSP, our 9th generation DSP, has enabled extraordinary reaches using even smaller QSFP-DD and OSFP form-factor modules. Acacia’s 400G ultra long-haul modules support distances beyond 3000km, and 800G ZR+ enhanced performance modules support distances beyond 1000km.

Leadership in Industry Standards
To ensure a thriving marketplace for coherent pluggable optics, many companies participate in MSA initiatives and/or in industry standards groups to agree on module product requirements to ensure a multi-vendor supply, interoperability, and economies of scale. Acacia has been a key driver and participant in many of these organizations. In the 400G pluggable generation, Acacia served as editor for the 400ZR Implementation Agreement in OIF. Acacia was also a founding member of the OpenZR+ MSA, which provided requirements for enhanced performance 400ZR+ with multi-vendor interoperability. Market adoption of 400ZR and OpenZR+ solutions has been the fastest of any coherent technology in history.

In the 800G pluggables generation, Acacia has continued its leadership in industry standards activities. As a key contributor along with other coherent suppliers, Acacia introduced the first standardized, interoperable probabilistic constellation shaping (PCS) mode for 800G ZR+ enhanced performance, enabling network operators to expand their 800G reaches beyond 1000km. The company is also playing a key role in driving 1600ZR/ZR+ industry agreements.

With the increasing demand for the current generations of 400G and 800G coherent pluggable modules, several companies have jumped on the bandwagon to participate in this optical renaissance. These 400G and 800G QSFP-DD and OSFP form-factor modules, capable of plugging directly into switch and router ports, have invigorated architectures using router-based coherent optics as a networking solution to reduce overall total cost of ownership.

Silicon-based Designs, Advanced DSPs, and High-Density Packaging
The key to Acacia becoming a market leader in coherent optics stems from our ongoing focus on core technology. For example, an important factor that provided many advantages in low power and high-density packaging was Acacia’s adoption of silicon photonics for high-speed coherent transmission. Silicon does not require temperature stabilization, which results in lower power implementation. In addition, silicon photonics can leverage high-volume CMOS packaging techniques to reduce overall cost. Our design and development approach has always been to utilize silicon photonics for both MSA pluggable modules as well as performance-optimized modules, providing design and operational efficiencies as well as rapid time to market. Following Acacia’s lead, we are now seeing multiple suppliers also embracing silicon photonics technology in their coherent pluggable modules because of the benefits it has delivered to the industry.

Another key technology advantage that sets Acacia apart is its coherent DSP design and capabilities. Innovation in this area has been critical for mitigating impairments such as dispersion and non-linear effects incurred during transmission over optical fiber. Acacia’s balanced and power-efficient approach to DSP development ensures robust algorithms while minimizing power consumption. As CMOS node sizes continue to shrink, relatively complex techniques for improving performance can be introduced into very compact MSA pluggable modules. The improved performance is enabling network operators to expand the types of architectures they can deploy using these pluggables, resulting in the ability to offer a wider range of network applications.

Driving innovation to meet increasing transmission baud rates meant having in-house cross-disciplinary teams to ensure the best designs. We were able to achieve this while also delivering a compact low power consumption device. Acacia’s 3D Siliconization enabled this high-density level of packaging with high signal integrity and allowed us to also leverage volume manufacturable processes used in the electronics industry. This approach enabled faster scaling of manufacturing and improved reliability.

15+ Years of Industry Firsts and Design Expertise
Technology leadership is a culmination of many factors, not just having the expertise in the technology itself. It is comprised of how the technology is developed and quickly brought to market with the capability of volume while maintaining quality, and with the goal of meeting customer requirements. Many factors have contributed to Acacia’s technology leadership, including an engineering approach based on in-house design expertise across key disciplines, leveraging years of experience, building on the successes of previous generations as well as adjacent industry innovations and manufacturing processes, and participating in industry standards. This powerful combination created many industry firsts over the last 15+ years since Acacia was founded.

Figure 3. Acacia has achieved many industry firsts for coherent pluggable modules since its founding.

Acacia has introduced coherent module products, including its recent Delphi-based pluggable modules, leveraging nine generations of DSP ASICs, and holds an industry leadership position when it comes to innovative design, quality, and manufacturability of coherent pluggable modules especially as baud rates increase while size and power consumption requirements decrease. Acacia continues to demonstrate its technology leadership as we move to higher bandwidth applications such as AI, as well as customer requirements for higher speed solutions such as 1600G MSA pluggable modules.

Market Adoption and Use Case Expansion
Coherent pluggables have seen remarkable market adoption, with 400G ZR/ZR+ coherent optics becoming the most widely adopted coherent technology in history. While some in the industry predicted early on that 400ZR pluggables would only be a small portion of the coherent market, Acacia has been laser focused on bringing these solutions to its customers to deliver the architectural change needed to transform networks to meet ever-growing bandwidth demands. As a result, Acacia is a market leader in shipments of 400G+ coherent pluggables.

During 2024, Acacia expanded this market-leading portfolio with the introduction of 800ZR and 800G ZR+ pluggables with Interoperable PCS in QSFP-DD and OSFP form factors. These solutions have already been proven in field trials, with Colt being the first provider to successfully trial enhanced performance 800G ZR+ coherent pluggable optics. These 800G router-based optics provide the capability to double Colt’s packet core capacity per link while reducing power per bit by 33.3%.

Acacia also introduced a 400G Ultra Long Haul QSFP-DD module for expanding 400G applications from DCI/metro to long haul applications. The 400G UHL has been proven in field trials, with Arelion announcing that it completed a live network field trial on its route from Chicago to Denver that demonstrated successful IP transmission at a spectrum of 112.5 gigahertz over 2,253 kilometers, with healthy margins, providing longer transmission distances and greater cost savings than currently deployed transponders. Acacia’s 400G ULH pluggables enable Arelion to reduce CAPEX by 35 percent and OPEX costs by 84 percent when expanding its network, providing wider reach with high capacities that support customers’ AI/ML and cloud applications.

The Terabit Era is Going Strong
According to Cignal AI, in 2025, 1.2T+ performance optimized solutions will contribute significantly to bandwidth growth as those solutions continue to be introduced. Acacia’s 1.2T Coherent Interconnect Module 8 (CIM 8), powered by the Jannu DSP, has proven its outstanding performance with multiple record-breaking field trials across Metro, Long Haul and Subsea with Microsoft, Verizon, Windstream, and others. Acacia is also engaged with multiple webscale customers and expects CIM 8 to continue its ramp in 2025.

Looking Ahead…..Artificial Intelligence, 1600ZR/ZR+ and More!
AI certainly was a hot topic during OFC and ECOC in 2024 and Acacia believes it will continue to be in 2025. While a year ago, the industry was focused on single-site AI clusters, we will begin to hear more about methods for distributing AI training over different locations in 2025. And while there is still a lot of work to happen in this area, the optics industry is well positioned to be at the forefront of this evolution since higher speed optical interconnections will be necessary to mitigate bandwidth constraints within an AI networks. Read this recent blog for more information on future proofing your network for AI.

Router-based Coherent Optics
The proliferation of router-based coherent optics is paving the way to a converged IP+Optical network architecture. The benefits in 2024 were clear. Infrastructure provider Colt Technology Services reported a stunning 97% energy savings, and Arelion saved 64% in CapEx and 76% in OpEx. In 2025, we can expect to see more providers leveraging this architecture to achieve increased capacity, reduced energy consumption, and lowered network costs, complexity, and footprint.

1600ZR and 1600ZR+ Standards Agreement
The OIF launched efforts last year on 1.6T coherent optical interconnect solutions and is making progress towards interoperable 1600ZR and 1600ZR+ implementation agreements.  In 2025, the industry will be looking at ways to advance this migration using advanced technologies for high baud rate modulation and smaller CMOS nodes.

Looking to the Future
Acacia is looking forward to another year of industry-wide innovation. Helping to solve some of our customers’ most challenging problems has been part of our DNA since Acacia was founded in 2009 and as you can see from this historic timeline, we have continued to set new benchmarks for performance and efficiency. These innovations are not only addressing the current demands of high capacity networks, but are also paving the way for future growth and scalability. Key to this success has been listening to our customers and designing the products and features that they need to be successful. We look forward to continuing that tradition.

With the initial adoption of 400G coherent pluggables for IP-over-DWDM networks being driven by router interconnects, these pluggable modules based on coherent technology have been referred to as router-based coherent optics. There are now more than 200 network operators that have embraced this cost-saving paradigm.

Figure 1. Router-based coherent optics provide cost savings.

400G Coherent Modules and Open Line Systems Led the Way
As previously mentioned, the introduction of 400G interoperable coherent MSA modules that plug directly into router ports helped accelerate network operator adoption of router-based coherent optics, enabling high-capacity optical connections within a metro reach network without traditional transponder hardware. Two different mechanical form-factors for these 400G modules, QSFP-DD and OSFP, were introduced to the market, with the former being the primary form-factor being shipped today for 400G, matching the widely adopted host platform QSFP-DD slots.

The disaggregation of optical line systems has also helped progress the adoption of router-based coherent optics. These open line systems enable the insertion of wavelength transmission from router-based coherent MSA pluggable modules rather than from transponders sold by the same line system vendor. Many of the recently deployed networks utilizing router-based optical modules have been over these open line systems. In fact, approximately 70% of the above mentioned 200 end-users were utilizing an open line system.

In addition, the introduction of 400G coherent modules with high transmit optical power, such as Acacia’s Bright 400ZR+ module, helped accelerate service provider adoption because higher transmit power helps to avoid performance penalties when connecting to typical brownfield ROADM architectures. Modules such as the Bright 400ZR+ also include a transmitter tunable optical filter (TOF) to minimize adjacent channel interference that could impact performance, especially if colorless ROADMs are present in the network.

An ongoing challenge that the industry is making progress with is the ability for seamless management of coherent MSA modules. Industry groups such as the Optical Internetworking Forum (OIF) have made great progress to address this challenge, with the OIF driving the Common Management Interface Specification (CMIS). This effort continues to be an area of industry focus to further lower the adoption barrier of router-based optics.

Continuing the Momentum of Router-based Coherent Optics
To continue the adoption of router-based coherent optics, expanding interoperable MSA pluggable module capabilities were required to address network operator use cases such as long haul and ultra long haul reaches as well as a migration from 400G links to 800G links. Thanks to recent advances in coherent technology, these capabilities have been recently introduced.

400G ultra-long-haul (ULH) modules leveraging Class 3 (~120+ Gbaud data) rate technology enables the reach capability of 400G to extend from metro/regional reaches to ultra long-haul reaches, reducing the barrier for network operators to deploy router-based coherent optics in virtually any network application. Arelion recently announced a successful trial using Acacia’s Delphi-DSP based 400G ULH modules over 2,253km with margin, enabling a 35% reduction in CAPEX and 84% reduction in OPEX.

To take advantage of the latest generational increase in switch/router chip capacity resulting in I/O ports transitioning from 400G to 800G speeds, the same Class 3 generation coherent technology support 800G interoperable coherent MSA modules that plug directly into host platforms. This enables network operators who have already embraced an IP-over-DWDM architecture using router-based coherent optics at 400G to now migrate to 800G. For example, Colt recently announced that it is the first provider to successfully trial enhanced performance 800G ZR+ coherent pluggable optics, in their Cisco 8000 series router ports, in its production network. These 800G router-based coherent optics provide the capability to double Colt’s packet core capacity per link while reducing power per bit by 33.3%.

While the initial adoption of router-based coherent optics for deploying an IP-over-DWDM network were from hyperscalers and service providers, the momentum of adoption has expanded to research and education networks, enterprise networks, and many other network operators looking to optimize total cost of ownership. And the application is not limited to using coherent pluggable optics in routers, but also in network switches for fabric extension requiring an interconnect to a distant site.

Acacia’s Interoperable Modules Enabling the Future
Acacia is enabling the adoption of 400G and 800G IP-over-DWDM architectures with router-based coherent optics. The latest generation of MSA pluggable modules include 800ZR and 800G ZR+ variants as well as 400G ULH for ultra-long-haul reaches. These are all powered by Acacia’s 9th generation Delphi DSP ASIC and 130+Gbaud high-speed silicon photonic PIC technology enabling a low-power industry standard based solution. The 800G ZR+ module also includes the industry’s first standardized interoperable probability constellation shaped (PCS) mode. In addition, Acacia is a leading supplier of 400ZR and OpenZR+ compliant modules including high Tx power Bright modules for 400G based metro/regional IP-over-DWDM network.

Enabled by coherent pluggable modules, the adoption of IP-over-DWDM using router-based coherent optics continues to grow, providing significant reduction in TCO for network operators.

The Path Towards Robust, Interoperable 1600ZR/ZR+ Interfaces
With 200G per lane electrical PAM4 solutions recently introduced, network operators now have a path towards supporting 1600G host router I/O ports by using eight parallel electrical lanes (Figure 1). Similar to 400G and 800G generations, this is a key motivator in developing coherent pluggable modules to be plugged into these 1600G router ports for inter-data center optical links. However, along with the progress on the host interface side, there is still much work to be done to ensure a technically feasible and robust interoperable design for 1600ZR/ZR+ coherent pluggable modules.

Figure 1. Simple illustration of how advances in achieving 200G PAM4 can be leveraged for 1600ZR/ZR+ coherent optical transmission.

OIF Defining Both 1600ZR and 1600ZR+ Standards
Unlike previous coherent standardization efforts at 400G and 800G, in which enhanced “ZR+” performance links were defined outside of OIF, the OIF has launched initiatives to define both 1600ZR and 1600ZR+. Having both efforts occurring simultaneously enables the OIF to make decisions with both 1600ZR and 1600ZR+ in the same scope of discussions. This helps keep the two implementations as aligned as possible, which is beneficial for the industry considering the large investments of technology required. The focus of these investments includes advanced CMOS nodes to maintain low power consumption within the envelope of QSFP-DD and OSFP form-factor requirements, and advanced designs in high-speed RF/mixed-signal as the modulation approaches the Class 4 240Gbaud range (Figure 2).

Figure 2. Charting the course towards Class 4 baud rate standardization efforts.

As we saw in both the 400G and 800G generations, the foundation of 16QAM (4 bits/symbol) modulation was adopted and this is likely to also happen with the 1600G generation. For 1600G transmission, 16QAM modulation implies ~236+Gbaud data rate operation.

In addition to modulation order, the type of forward error correction (FEC) has also been a key parameter that required industry agreement. At 400G, the OIF adopted concatenated FEC (CFEC) as the 400ZR FEC and OpenZR+ MSA adopted oFEC (a high-performance FEC) for 400G ZR+. At 800G, the OIF decided to adopt oFEC for ZR, aligning it with ZR+ modes. To provide an enhanced performance mode beyond 800ZR, OpenROADM MSA defined an interoperable PCS for 800G ZR+ (Figure 3). It is likely that oFEC will be similarly adopted for both 1600ZR and combined with some interoperable PCS for 1600ZR+ modes.

Figure 3. 400G to 800G evolution of ZR vs. ZR+ implementations; how will 1600G ZR vs. ZR+ implementations be different?

What Will Be the Industry Consensus for 1600ZR/ZR+?
Every new generation of speeds-and-feeds encounters challenges around industry consensus and technology achievements that push the envelope – and 1600ZR/ZR+ is no different. There is currently great momentum driving these efforts forward, especially in anticipation of advances in generative AI that are pushing optical interconnect needs to higher bandwidths. Evidence of this momentum is apparent by other industry efforts beyond the OIF that are currently active. In addition to the OIF 1600ZR/ZR+ efforts, the IEEE has also begun working on 1.6TbE electrical and optical interface standards within the IEEE 802.3dj working group, anticipated to be ready by the second half of 2026.

In light of this progress, the question is “how does the industry reach consensus for 1600ZR/ZR+?” We eagerly await the outcome.

The Optical Internetworking Forum (OIF) kicked off the 400G MSA pluggable generation with development of the 400ZR implementation agreement enabling point-to-point amplified links up to 120km operating at 60+Gbaud data rates. Around the same timeframe, the OpenROADM MSA defined 400G interfaces for ROADM networks and extended reaches; the OpenZR+ MSA leveraged these higher performance interfaces to enable interoperable enhanced performance links for 400G pluggable modules (Figure 1).

The introduction of high-transmit optical power (>0dBm) ZR+ modules such as Acacia’s Bright 400ZR+ module further expanded the 400G MSA pluggable space to include brownfield ROADM network architectures (with existing transponder channels ~0dBm). Driven by increasing bandwidth demands from applications such as AI, network operators are now looking towards a new generation of MSA pluggable products that further expand applicable networking scenarios that operators can leverage to scale and meet these demands.

How Industry Standards Benefit MSA Pluggable Module Adoption
The latest array of MSA pluggable products introduces a new set of capabilities that network operators can utilize to increase capacity and extend reach. These products provide the ability to deploy 800G with ZR, ZR+, and high-transmit optical power capabilities, as well as extending the capabilities of existing 400G router interfaces to support ultra-long-haul (ULH) reach capabilities. This new generation of modules continues to leverage industry standardization while also borrowing capabilities from performance-optimized coherent solutions. These capabilities include high-baud rate transmission allowing for a doubling of baud rates from the previous Class 2 (~60+Gbaud range) generation to Class 3 (~120+Gbaud range) baud rates, the use of probabilistic constellation shaping (PCS) for enhanced transmission performance, and L-band support for spectrum range expansion.

Figure 1.  Interoperability approaches at 400G vs. 800G.

Industry standardization of coherent solutions plays a key role in enabling economies of scale. Users of 400G coherent MSA pluggable modules such as 400ZR/ZR+ have benefited from the efforts of OIF, OpenZR+ MSA, and OpenROADM MSA to provide industry agreements on module specifications resulting in a diverse supply base. We have seen similar efforts to garner industry standardization as users transition to 800G MSA pluggables. There are three main elements that differentiate 800G relative to 400G and are adapted from previously developed performance-optimized solutions.

1. Interop PCS for Enhanced Performance
   A key difference between 400G and 800G interoperability approaches for an enhanced performance “ZR+” is that instead of using enhanced performance forward error correction, oFEC, to provide improved 400G performance, 800G uses industry standard interoperable probabilistic constellation shaping (PCS) for enhancing performance. PCS is a transmission shaping technique that provides additional link performance beyond traditional transmission modes such as 16QAM. Industry standardization of an interoperable PCS transmission shaping function, once relegated to proprietary performance-optimized transponder platforms including those for submarine applications, is a tremendous leap forward in the progress of MSA pluggable module capabilities. Multi-vendor 800G module supply chain diversity from a DSP ASIC perspective is possible when the 800G ZR+ performance enhancement mode utilizes the industry standard interoperable PCS mode.
2. High Baud Rate Design
   PCS is not the only technology that has been adapted from performance-optimized solutions for MSA pluggables. 800G as well as a 400G ULH pluggable solutions require a high-baud rate design operating in the Class 3 ~120+ Gbaud data rate range. Acacia’s performance-optimized CIM 8 module capable of 140Gbaud speeds has already proven that its deployed technology far exceeds the requirement for the new generation of MSA pluggables. Operation at these high baud rates benefits heavily from the advanced integration and RF signal optimization techniques that Acacia introduced in our 400G MSA pluggable product family.

Figure 2.  Tightly integrated components enable 120+Gbaud data-rate capabilities.

3. C & L Band Support
A third element of the latest 800G MSA pluggable generation that is borrowed from performance-optimized designs is the capability to transmit in the L-band wavelength range, in addition to the traditional C-band DWDM range. By adding L-band supporting infrastructure to a network, the network capacity is approximately doubled. Network operators now have an option beyond utilizing a transponder platform if they wish to use L-band expansion to increase network capacity.

Figure 3.  New generation of coherent MSA pluggable modules to take advantage of L-Band transmission window, adding to existing C-Band support.

Pluggable Interoperable Interfaces are Driving Adoption of 800G Modules
Acacia’s latest family of coherent solutions are powered by its 9th generation DSP ASIC called Delphi. These modules include support for OIF 800ZR, interoperable 800G ZR+ using the OpenROADM interop PCS mode, and 400G ULH for ultra-long-haul reaches. These modules utilize Acacia’s 3D Siliconization providing a highly integrated design enabling high-baud rate modulation. With support for QSFP-DD and OSFP form factors, as well as >+1dBm transmit optical power and L-band support, Acacia’s Delphi generation of products leverage the deployment successes of our performance-optimized CIM 8 module to provide MSA pluggable products that offer increased capacity and longer reaches.

Figure 4.  Acacia’s latest generation of MSA pluggable 800G and 400G ULH modules.

Similar to the successful path we saw 400G pluggables experience, these modules are delivering the performance and interoperability that is critical for driving economies of scale and widespread adoption. With data center bandwidth continuing to grow rapidly, fueled by emerging new applications such as AI, these high-performance pluggable modules are on track to become an important tool for network operators to cost-efficiently scale their networks to meet this surging demand.

See Us at ECOC 2024!
Acacia is excited to be participating in the OIF interoperability demo at ECOC 2024 showcasing both its 400G and 800G pluggables; demos will take place in the OIF booth #B83. Acacia will also be demonstrating the Interoperable 800G ZR+ module in our meeting room at ECOC. Click here to set up a meeting.

We hope to see you in Frankfurt!

AIs Effect on Optical Transport Network
While the near-term focus on high-capacity interconnects for AI applications has been on short reach connections within AI clusters, the industry is already seeing bandwidth requirements begin to increase, requiring additional coherent connectivity between datacenters supporting AI. And while there is general agreement that the resulting bandwidth demand from AI applications translates to increased traffic across the network, the industry is in the early stages of understanding how specific segments of the network are affected.  Coherent optical interconnects for high-capacity transport beyond the data center already provide performance-optimized transponder solutions at 1.2T per wavelength as well as 400G router-to-router wavelengths moving to 800G using MSA pluggable modules. As discussed in this recent blog, we believe this technology can continue to play a role with expanding traffic demands in the metro, data center interconnect, long haul, and beyond.

To learn more about how AI will impact optical networks, stop by the ECOC workshop on 9/22 at 14:00 CET and listen to Acacia’s Timo Pfau give a talk titled “How Will AI Affect Future Transmission Systems?”

Interoperable 400G and 800G Demonstrations
Following the explosive growth of 400G pluggables, next-generation switch/router ASICs are now being introduced with 800G I/O port speeds, creating the need for 800G optical interfaces for data center interconnect (DCI.) As we will hear at this year’s ECOC, these pluggables are following the same 400G MSA pluggable cycle, with multiple standardization bodies such as OIF, Open ROADM and the IEEE working in parallel to provide industry standards for 800G optical transmission and 800 Gigabit Ethernet (GbE).  Interoperability will be key for driving the economies of scale that can enable these modules to experience similar levels of success as 400G while also helping to define the path to 1.6T pluggables in the future.

As a leader in 400G coherent pluggable shipments, Acacia is excited to be participating in the OIF interoperability demo showcasing both its 400G and  800G pluggables. The demos will take place in the OIF booth #B83 and will spotlight interoperability innovations in 800ZR, 400ZR and multi-span optics and Common Management Interface Specification (CMIS) – all pivotal for shaping the next decade of industry standards.

Announced earlier this year, Acacia’s newest portfolio of silicon-based 800G coherent pluggables are powered by Delphi, Acacia’s 9^th generation DSP.  These modules have been designed to provide double the connectivity speed from 400G to 800G to support DCI upgrades, and the high baud rate design has already been proven in Acacia’s deployed CIM 8 module. In addition, the Interoperable 800G ZR+ modules in this family were the first in the industry to support the OpenROADM specifications that include interoperable PCS transmission capability, which expands the market for 800G pluggables beyond simple DCI into regional and even long-haul networks. Acacia will be hosting private demonstrations of its Interoperable 800G ZR+ module in its private meeting room.

To meet customer needs and drive future adoption, it’s important for 800G coherent MSA pluggables to have several key features and capabilities including optical transmission, client traffic, low power consumption, and interoperable module management.  One of the key functional requirements that operators are looking for is Coherent-CMIS compliance. This feature provides a well-defined mechanism to initialize and manage optical and copper modules in a standard way, while still providing the capability to provide custom functionality. This commonality makes integration into different host platforms easier for both the host and module vendors.

To learn more about CMIS compliance, come see Acacia’s Doug Cattarusa give an ECOC Market Focus presentation titled “CMIS:  Is Plug-and-Play Possible?” on Monday, 9/23 at 14:40 CET.

Expanded Use Cases for Router-based Optics
The introduction of 400G coherent pluggable optics in metro reach applications enabled the collapsing of network layers, converging the optical transport and IP layers. A recent report from Cignal AI predicts IP-over-DWDM port deployments to top 700k in 2027. Recent advances in coherent technology allowed for the expansion of coherent and interoperable MSA pluggable module capabilities into long haul and ultra long haul reaches as well as migration from 400G links to 800G links. Acacia’s 400G ultra-long-haul (ULH) modules leveraging 130+ Gbaud data rate technology enable the reach capability of 400G to extend from metro/regional reaches to ultra long-haul reaches, reducing the barrier for network operators to deploy IP-over DWDM adoption for network operators desiring IP-over-DWDM beyond metro/regional applications.

OIF Takes the Lead in Defining 1600ZR and 1600ZR+ Standards
The OIF launched efforts last year on 1.6T coherent optical interconnect solutions and is already making progress towards interoperable 1600ZR and 1600ZR+ implementation agreements.  While there is still much work to be done in the standards bodies around 1.6T, we expect this year’s ECOC to be a sounding board for the industry to discuss solutions for some of the key challenges around this transition. This includes the use of advanced technologies for high baud rate modulation and smaller CMOS nodes not yet supported in volume for high baud rate modulation, power consumption, options for single or dual optical carrier, internal optical amplification requirements, and backwards compatibility to 800ZR/ZR+ and 400ZR/ZR+ interfaces.

See Us at ECOC 2024!
If you are attending ECOC this year and want to connect, we’d welcome the opportunity to meet with you. Click here to set up a meeting. We hope to see you in Frankfurt!

While the near-term focus has been on how AI will affect the technology around short reach interconnects, there will certainly be an impact to interconnections beyond the AI clusters and beyond the AI data center, in hyperscaler networks.

The question is:  beyond the short-distance high-bandwidth interconnections, how would AI traffic impact the optical transport environment beyond the intra-building network and into the metro, long haul, and longer reach applications where optical coherent transmission is heavily utilized?

Figure 1.  Sales Forecast for Ethernet Optical Transceivers for AI Clusters (July 2024 LightCounting Newsletter, “A Soft Landing for AI Optics?”).

Effect of Past Applications on the Transport Network
Bandwidth-intensive computing is attributed to both AI training as well as AI inference, where inference refers to the post-training process in which the model is “ready for the world,” creating an inference-based output from input data using what it learned during the training process. In addition to AI training’s requirements for a large amount of computing power and a large number of high-bandwidth, short connections, there are also foreseen bandwidth requirements beyond the AI data center. To understand how network traffic patterns may evolve beyond the AI data center, let’s review some examples of how the wider transport network was affected by past growth of various applications. Although these applications may not strongly match the effect of AI application traffic, it can provide some insight into the effects that the growth of AI applications may have on optical transport, and thus on the growth of coherent technology.

If we look at search applications, the AI training process is generally analogous to a search engine’s crawling bots combing the internet to gather data to be indexed (AI training being much more computationally intensive). The AI inference process is analogous to the search engine being queried by the end user with results made available for user retrieval with minimal latency. While the required transport bandwidth for search bots and user queries are minimal compared to higher bandwidth applications, the cumulative effect of the search-related traffic is part of the contribution to overall transport traffic, including bandwidth from regional/local caching to minimize latency, as well as usage from subsequent traffic created by acting upon search results.

Understanding how network traffic was affected by the growth of video content delivery is another example that can inform potential AI network transport traffic patterns. A main concern resulting from video content distribution was the burden imposed on the network in delivering the content (especially high-resolution video) to the end-user. To address this concern, content caching, where higher demand content was cached closer to the end-user, was implemented to reduce overall network traffic from the distribution source to the end-user, as well as reduce latency. While it is too early to predict how much network traffic would increase due to expansive queries to and responses from AI inference applications, the challenge is to ensure the latency for this access is minimal. One could see an analogy of content caching to edge computing where the AI inference model is closer to the user with increased transport bandwidth required for these edge computing sites. However, the challenge would be to understand how this would affect the efficiency of the inference function.

Turning to cloud computing for insights on traffic patterns, the rise of (multi-) cloud and computing resulted in intra and inter-datacenter traffic (a.k.a. east-west traffic) increasing as workloads traversed across the datacenter environment. There’s a similar potential rise in this type of traffic with AI as data for training could be dispersed among multiple sites of clusters as well as inference models being distributed to physically diverse sites to reduce latency to end users.

For any of these previous examples, as the demand of these applications increases, the transport bandwidth requirements would also increase from not only the target data (e.g., search results, video), but also from overhead or intra datacenter traffic to support these applications (e.g., content caching, cloud computing, backend overhead). Traffic behavior for aggregating AI training content as well as the distribution of AI inference models and its results may be similar to the traffic patterns of these previous applications, applying pressure to network operators to increase capacity for its data center interconnect, metro, and regional networks. Long haul and subsea networks may also experience a need to expand to meet the demands of AI-related traffic.

Figure 2.  A scenario in which the network fabric physically expands due to facility power constraints, requiring high-capacity optical interconnections.

The Balance of Power and Latency
While the application examples above are related to how the AI application itself may affect bandwidth growth, what is becoming apparent is the power requirements to run AI clusters and data centers are significant. In the past, as the demand for cloud services grew, the need for large-scale data centers to have access to localized inexpensive power sources helped to drive the location selection for large data centers. However, power facility/availability constraints helped drive the adoption of physically distributed architectures, which then relied on high-capacity transport interconnects between data centers to maintain the desired network architecture (Figure 2). We anticipate a similar situation with AI buildouts requiring distributed facilities to address power constraints with potential trade-offs of reduced efficiencies for both AI training and inference. The distributed network would then rely on high-capacity interconnect transport using coherent transmission to extend the AI network fabric. Unlike cloud applications, physical expansion of the network fabric for AI applications has a different set of challenges due to compute and latency requirements for both training and inference.

Figure 3. Extremely low latency is required within the AI cluster to expeditiously process incoming datasets during the training mode. Since datasets are collected before being fed into the training cluster, the process of collecting these datasets may not be as latency sensitive.

As we plan for AI buildouts, one common question is how the physical extension of an AI networking fabric may affect AI functions. While geographic distribution of AI training is not ideal, facility power constraints are certain to lead to a growing adoption of distributed AI training techniques that attempts to mitigate introduced latency effects. As part of the training process, sourcing datasets feeding into the training cluster may not be latency sensitive and would not be as impacted by physical network extension (Figure 3). After training, when the inference model is complete, the goal is to minimize the latency between the user query to the inference model and the transmitted results to the user (Figure 4). The latency is a combination of the complexity of the query as well as the number of “hops” between the inference model and the user. Latency reduction when accessing the inference model, as well as methods to effectively distribute both the training and the inference function beyond a centralized architecture to address single-site power constraints, are ongoing discussions within the industry.

Whether driven by power constraints, dataset sourcing, or inference response efficiency, the sheer growth of AI applications will drive network traffic growth beyond AI cluster sites towards the wider network requiring high-capacity interconnects.

Figure 4.  Minimizing latency for AI inference is a key objective.

Trading off power requirements versus access to inexpensive and abundant power versus latency is familiar territory when it comes to bandwidth intensive applications. The outcome that optimizes these trade-offs is application dependent and can even be deployment-by-deployment dependent. We continue to watch the evolving AI space to see how these network architecture trade-offs will play out, with the impact of how the transport network is designed. High-capacity coherent transport can certainly influence these trade-offs. And as we have already seen, by using coherent high-capacity transport cloud architectures, networks were able to physically expand to alleviate power source constraints by provided fat-pipe links between sites. We anticipate a similar scenario with expanding AI network architectures.

The Ripple Effect
While the near-term focus on high-capacity interconnects for AI applications has been on short reach connections within AI clusters, we are already seeing bandwidth requirements begin to increase, requiring additional coherent connectivity between datacenters supporting AI. And while there is general agreement that the resulting bandwidth demand from AI applications translates to increased traffic across the network, we are at the early stages in understanding how specific segments of the network are affected. Coherent optical interconnects for high-capacity transport beyond the data center already provide performance-optimized transponder solutions at 1.2T per wavelength as well as 400G router-to-router wavelengths moving to 800G using MSA pluggable modules. This technology will continue to play a role in the transport solution supporting AI applications whether the expanding traffic is in the metro portion, data center interconnects, long haul, or beyond.

Increasing Baud Rates Delivers Twice the Capacity and Longer Reaches
History has shown that increasing baud rate is an efficient way to enable more cost-effective optical networks by reducing the number of optics required to support a given transmission capacity. As highlighted in the graphic below, by doubling the baud rate between successive generations, Class 3 products such as Acacia’s CIM 8 1.2T module supports twice the capacity per carrier over greater reaches than Class 2 solutions. This approach provides a simple, scalable path that supports higher capacity per carrier over the reaches needed for existing network architectures.

Network operators leveraging Class 3 140Gbaud solutions for DCI, metropolitan, long-haul, and subsea applications can maximize their network coverage across a wider distance than ever before. This results in the capability to transmit native 400G client traffic over virtually any network application, delivering 3x400G over 1.2Tbps per carrier capacity for high-capacity DCI interfaces, 2x400G over 800Gbps per carrier capacity for most optical links using 4 bits/symbol (~16QAM) modulation and 2 bits/symbol (400G QPSK) over ultra-long-haul and subsea distances.

Acacia’s Dayou Qian, Product Line Manager, will discuss the migration to high baud rate 400G QPSK during a live panel presentation titled Class 3 Optical Transport Systems: from DCI to Subsea at OptiNet China on June 20.

Learn More about Acacia’s Performance Optimized Solutions
As the first coherent module on the market to break through the terabit threshold, Acacia’s 1.2T CIM 8 module has proven its outstanding performance with multiple record breaking field trials across a wide range of applications. Visit this link to learn more.