October 2024 - Dell'Oro Group

The 2024 OCP Global Summit theme was “From Ideas to Impact,” but it could have been “AI Ideas to AI Impact.” Accelerated computing infrastructure was front and center starting with the keynote, on the exhibition hall floor, and in the breakout sessions. Hyperscalers and the ecosystem of suppliers that support them were eager to share what they’ve been working on to bring accelerated computing infrastructure and AI workloads to the market, at scale. As you may expect with anything AI related, it drew a crowd – Over 7000 attendees participated in the event in 2024, a significant increase from ~4500 last year. Throughout the crowds, sessions, and expo hall, three key themes stood out to me: Power and cooling designs for NVIDIA GB200 NVL Racks, an explosion of interest in liquid cooling, and sustainability’s presence among the AI backdrop.

Powering and Cooling NVIDIA GB200 NVL Racks

It’s well known that accelerated computing infrastructure significantly increases rack power densities. This has posed a significant challenge for traditional data center designs, where compute and physical infrastructure are developed and deployed in relative isolation. Deploying accelerated computing infrastructure has forced a rethink, where these boundaries are removed to create an optimized end-to-end system to support next generation “AI factories” at scale. The data center industry is acutely aware this applies to power and cooling, with notable announcements and OCP contributions from industry leaders in how they are addressing these challenges:

Meta kicked off the keynote by announcing Catalina, a rack-scale infrastructure design based on NVIDIA GB200 compute nodes. This design increased the power requirements from 12 – 18 kW/rack to 140 kW/system. To no surprise, Catalina utilizes liquid cooling.
NVIDIA contributed (open-sourced) elements of its GB200 NVL72 design, including a powerful 1400-amp bus bar for distributing power in the rack, and many liquid cooling contributions related to the manifold, blind mating, and flow rates. Lastly, NVIDIA recognized a new ecosystem of partners focused on the power and cooling infrastructure, highlighting Vertiv’s GB200 NVL72 reference architecture, which enables faster time to deployment, utilizes less space, and increases cooling energy efficiency.
Microsoft emphasized the need for liquid cooling for AI accelerators, noting retrofitting challenges in facilities without a chilled water loop. In response, they designed and contributed a custom liquid cooling heat exchanger, which leverages legacy air-based data center heat rejection. This is what I would refer to as air-assisted liquid cooling (AALC), more specifically, an air-assisted coolant distribution unit (CDU), which is becoming increasingly common in retrofitted accelerated computing deployments.
Microsoft also announced a collaborative power architecture effort with Meta, named Mt. Diablo based on a 400 Vdc disaggregated power rack, that will be contributed to the OCP soon. Google also highlighted the potential use of 400 Vdc for future accelerated computing infrastructure.

Data Center Liquid Cooling Takes Center Stage

Liquid cooling was among the most discussed topics at the summit, mentioned by nearly every keynote speaker in addition to dozens of breakout sessions dedicated to its growing use in compute, networking, and facility designs. This is justified from my perspective, as Dell’Oro Group previously highlighted liquid cooling as a technology going mainstream creating a $15 B market opportunity over the next five years. Furthermore, the ecosystem understands that not only is liquid cooling a growing market opportunity, but a critical technology to enable accelerated computing and the growth of AI workloads at scale.

There was not just liquid cooling talk, but partnerships and acquisitions leading up to and during the global summit that further cemented the critical role data center liquid cooling will play in industries’ future. This was highlighted in the following announcements:

Jabil acquired Mikros Technologies: Kicking off two weeks of big announcements, Jabil’s acquisition of Mikros brings together Mikros’s expertise in liquid cooling cold plate technology, engineering and design with Jabil’s manufacturing scale. This appears to position Mikros’s technology as a high-volume option for hyperscale end-users and the greater data center industry in the near future.
Jetcool announced facility CDU, Flex partnership: Jetcool, most known for their air-assisted liquid cooling infrastructure packaged in single servers, introduced a facility CDU (liquid-to-liquid) to keep pace with the market’s evolution towards purpose-built AI factories. The partnership brings together a technology specialist with a contract manufacturer to enable the coming scale needed to support hyperscale end-users and the greater data center industries’ liquid cooling needs.
Schneider Electric acquired Motivair: On the Summit’s final day, Schneider Electric announced its $1.13B acquisition of Motivair. This move, following prior partnerships and organic CDU developments, expands Schneider’s high-density cooling portfolio. This now gives Schneider a holistic power and cooling portfolio to support large-scale accelerated computing deployments, a capability previously exclusive to Vertiv, albeit at a high cost for Schneider.

Sustainability Takes a Back Seat but Is Still Very Much Part of the Conversation

While sustainability did not dominate the headlines, it remained a recurring theme throughout the summit. As AI growth drives massive infrastructure expansion, sustainability has become a critical consideration in data center designs. OCP’s CEO George Tchaparian characterized sustainability’s role alongside AI capex investments best, “Without sustainability, it’s not going to sustain.” Other highlights include:

OCP announced a new alliance with Net Zero Innovation Hub, an organization focused on net-zero data center innovation in Europe. Details on the alliance were sparse, but more details are expected to emerge on this partnership at the 2025 OCP EMEA Regional Summit.
Google shared a collaboration with Meta, Microsoft, and Amazon on green concrete. Most impressively, this collaboration began with a roadmap around the time of last year’s OCP Summit, which resulted in a proof-of-concept deployment in August 2024, reducing concrete emissions by ~40%.
A wide range of other sustainability topics were discussed. Improvements in cooling efficiency, water consumption, heat reuse, clean power, lifecycle assessment, and metrics to measure and track progress related to data center efficiency and sustainably were all prevalent.

Conclusion: Data Center Power and Cooling is Central to the Future of the Data Center Industry

The 2024 OCP Global Summit left me as confident as ever in the growing role data center power and cooling infrastructure has in the data center industry. It’s not only improvements to existing technologies but the adoption of new technologies and facility architectures that have emerged. The event’s theme, “From Ideas to Impact,” serves as a fitting reminder of how AI is reshaping the industry, with significant implications for the future. As we look ahead, the question isn’t just how data centers will power and cool AI workloads, but how they’ll do so sustainably, efficiently, and at an unprecedented scale.

Late last week, Vecima Networks announced that it was acquiring Falcon V, a Polish developer of access network orchestration software designed to facilitate the deployment of vendor-agnostic DOCSIS, fiber, and wireless networks. The acquisition will help accelerate Vecima’s Entra vCMTS product development and help the company build closer ties to Charter as the cable operator continues its Distributed Access Architecture (DAA) network transformation. The deal also helps to soothe the sting of Vecima’s unsuccessful bid for the cable assets of Casa Systems, despite establishing itself as the stalking horse bid in the auction.

Falcon V, which originated in 2018 as a joint venture between Liberty Global and equipment supplier Vector Group, received an investment from Charter and Liberty Global in 2021 to focus on developing SDN and NFV solutions to allow for the deployment of open DAA systems. At the time of the investment, Charter was focused on deploying Remote MACPHY technology, as opposed to Comcast and other operators, who were moving forward with Remote PHY. Falcon V was said to be working on software that could accelerate vendor interoperability and help Charter move more quickly in the direction of Flexible MAC Architecture (FMA), which offered the operator far more flexibility in where it could locate the MAC (Media access control) function, be it in nodes, hub sites, headends, or centralized data centers.

But in October 2022, Charter changed direction and moved away from Remote MACPHY toward Remote PHY. That strategic shift left many wondering whether Falcon V would still have a role to play in Charter’s transition to DAA. In actuality, nothing changed much for the software supplier, as it was still focused on developing orchestration software as well as an interop testing suite designed to ensure Charter could have a truly open, vendor-agnostic DAA network.

In March 2023, Charter announced that it had selected Harmonic as a vCMTS and Remote PHY Device (RPD) technology supplier while also selecting Vecima as a supplier of its ERM 3 RPDs, which can be installed in its EN 9000 Generic Access Platform (GAP) nodes, all clearly indicating its commitment to a multi-vendor deployment. Vecima had already been selected as the lead supplier of Remote OLTs (R-OLTs) in Charter’s RDOF network buildouts and is presumably a lead supplier of these platforms in potential non-RDOF deployments, as well.

In September 2023, Vecima also announced it had entered into a warrant agreement with Charter, providing Charter the opportunity to purchase up to 361, 050 shares of Vecima stock through 2031 at a strike price of C$17.09 per warrant. That translates into an agreement of roughly US$4.5M and is dependent on Charter achieving certain spending targets.

So, even before the Falcon V acquisition, the relationship between Vecima and Charter was already strong. The addition of Falcon V and its employee base extends that relationship further into the realms of vCMTS, software orchestration, and DAA interop testing.

An Answer to Charter’s Interop Issues?

Back in February 2024, Charter’s Chris Winfrey announced that the start of phase two of its network transformation—the phase focused on RPD and vCMTS deployments—would be delayed from the beginning of the year to late 2024, at best. The culprit? DAA equipment certification delays due to greater-than-expected challenges with interop testing. Though Winfrey didn’t provide specifics on the delays, Charter’s multi-vendor strategy is already ambitious, especially when the company continues to build out RDOF properties with R-OLTs and is also trying to roll out new nodes and amplifiers.

Thus, Vecima’s acquisition of Falcon V could very well have been pushed by Charter as a way to reduce the number of discrete vendors it has to coordinate with as it goes through the interop and homologation process. Charter has already made financial commitments to both vendors, so why not advocate for a marriage to help potentially speed up the DAA rollout process? The double-edged sword of DAA network rollout delays and subscriber losses is beginning to weigh heavily on Charter’s investors. So, anything that its vendor partners can do to solve those issues will certainly be welcomed by the operator.

Accelerating Vecima’s vCMTS Development

Beyond tightening its relationship with Charter, the addition of Falcon V’s products, as well as its software development teams will certainly help bring Vecima’s Entra vCMTS platform to market more quickly so that it can compete with Harmonic and Commscope. Though the Falcon V acquisition doesn’t completely make up for missing out on acquiring Casa’s cable assets, including its Axyom vCMTS and vBNG platforms, it does help to add pieces to what is an incredibly complex platform.

Vecima needs to accelerate the time to market of its Entra platform, especially at a customer like Charter, which has said it wants to move forward with a multi-vendor core, not just a multi-vendor PHY layer. While the details of just what a multi-vendor core might look like and how it will benefit Charter with all of the many balls it already has in the air, it certainly represents an opportunity for Vecima to position itself with a major operator that has plans beyond just the upgrade of its HFC network.

Charter likely similarly views the vCMTS as Comcast: As an edge compute platform that will ultimately enable services beyond those in the DOCSIS realm. The first workload after vCMTS is vBNG to support FTTH services and then perhaps an AGF (Access Gateway Function) workload to deliver converged fixed and mobile services over the existing HFC plant. Beyond that, perhaps a truly converged fixed and mobile core.

At last week’s SCTE TechExpo, Comcast announced that Rogers Communications, Canada’s largest cable operator, will be licensing Comcast’s broadband access network design. This includes DAA equipment, CPE, and other network monitoring and management elements. Now, Rogers syndicates Comcast’s broadband access architecture and components, in addition to the X1 video platform, which it has been licensing since January 2017.

This announcement was only a surprise in its timing, as rumors had been swirling for months, if not years, that Comcast would be syndicating its broadband network design and corresponding network elements to other cable operators. In fact, back in 2017, I wrote an article titled “Comcast’s Hands-on Approach to the Headend and Home,” where we speculated about Comcast moving in this direction. This was largely because it had already gained experience by syndicating its X1 video platform and aimed to “streamline the lengthy cycle of product definition, development, testing, homologation, and deployment. Comcast has signaled its intention to define the future access network in conjunction with its technology suppliers. Through the development of an RDK-like operating system designed to provide a standard reference point for the creation, provisioning, and management of broadband traffic and services.”

Virtualizing the Broadband Access Infrastructure

Further, the article argued that Comcast is working on virtualizing its broadband access infrastructure and all the software elements in order to create an operating system for DOCSIS networks and broadband services, in general. “Again, we believe the idea here is for Comcast to exert the same level of operational control over its broadband access network as it is doing in subscribers’ homes (with X1), by developing an access network operating system upon which all CCAP, optical node and optical line terminals (OLTs) will run. This operating system will extend the existing DOCSIS specification into a virtualized environment, providing specifics for the provisioning of broadband services across the entire access network. Additionally, it will incorporate APIs to tie into RDK-B and extend service provisioning into the home. This new system, called s RDK-A (RDK for Access) will allow Comcast and any of its licensees to move faster toward a more virtualized future.”

Comcast’s Partnership with Harmonic and CableOS

At the time the article was written, the focus was on software, particularly the CableOS vCMTS platform from Harmonic that Comcast had selected to serve as its primary platform moving forward. Back in September 2016, Harmonic announced a warrant agreement with Comcast allowing the operator to potentially acquire 7.8 million shares of Harmonic stock based on sales and deployment milestones of Harmonic’s CableOS product. At the time, the article noted that “the agreement with Harmonic is interesting because the company has not been a major supplier of CCAPs to Comcast. But Harmonic’s CableOS (now cOS) platform addresses the anticipated changes Comcast and other MSOs will see in their broadband access networks.” We speculated that “not much is known regarding the specific software elements that Comcast is looking to incorporate into its longer-term vision of an access network OS. If this is the case, it is a potentially huge win for Harmonic. It could mean the licensing of its CableOS software to other cable operators. For Comcast, this would ultimately mean more control, technologically and economically, over how broadband services are created and delivered from its network, which will be absolutely critical as broadband encompasses fixed, Wi-Fi, and mobile networks.”

We could not have predicted that Comcast’s licensing blueprint would go well beyond software and control plane functions to also encompass amplifiers, remote PHY devices, and machine learning tools all designed to help cable operators reduce their time to market and improve their overall network reliability.

Why Rogers is the Right Partner

Beyond having already been a long-time customer of Comcast’s X1 video platform and having recently signed a 10-year deal for ongoing access to Comcast’s video platform and CPE, Rogers Communications was the right partner at the right time for Comcast, for two reasons: First, the company has been occupied with its massive merger with Shaw Communications, Canada’s second-largest cable operator. The deal, which was first announced in 2021, officially closed in 2023. However, the hard work of bridging their networks and vendors is still ongoing, taking valuable network planning resources and personnel.

Second, back in July 2022, Rogers experienced a major network outage that impacted not only 12 million Rogers broadband and mobile customers, but also a number of ISPs with wholesale access to Rogers’ network. The outage, lasting from 15 hours to multiple days, resulted in Rogers having to give out approximately $150 million in customer credits. Rogers also developed a $10 billion plan to improve network reliability over three years to prevent, or at least minimize future outages.

Adopting Comcast’s broadband access network blueprint made sense for an operator under heavy pressure to prove to its subscribers and the Canadian Government that its network challenges were behind it. Now, Rogers can focus on marketing and selling its services, especially as it faces intensifying competition from Telus and Bell Canada, both of which are moving forward with major fiber overbuilding projects.

Which Access Technology is Right for Rogers?

Comcast is moving full-speed ahead with Full Duplex (FDX) DOCSIS 4.0. Rogers has already publicly communicated that it is testing FDX technology in its lab. However, is that the right choice for Rogers, and, does the licensing of Comcast’s access network blueprint mean it is on a path to FDX, as opposed to Extended Spectrum (ESD)?

Similar to Comcast, Rogers has both a node plus zero portion of their HFC networks, as well as a more traditional node plus five or six portions. We estimate that the node plus zero covers around 1 million homes, while the more traditional HFC plants 3 million homes. In that node plus zero portion, Rogers has deployed GPON but is in the process of upgrading to XGS-PON. Shaw also has a smaller fiber footprint, but instead of GPON, Shaw has deployed 10G EPON.

For the bulk of its HFC footprint, Rogers was previously rumored to be on the path toward deploying ESD using 1.8GHz amplifiers. However, given the new licensing arrangement with Comcast and the additional spectrum management tools the Comcast solution provides, Rogers could very well be considering deploying more FDX throughout its system. This is especially the case if the amplifiers are proven to work consistently in node plus 6 and even node plus eight environments. Both the Rogers and Shaw plants have long spans and larger amplifier cascades to deal with.

Though FDX amplifiers are presumed to be considerably more expensive than 1.8GHz amplifiers, it can be argued that the cost savings in not having to swap out taps, which is required in ESD deployments, makes FDX a wash on a per-home basis. Time will tell whether this is true or not.

Also, there is something to be said for the idea of being able to use the shared spectrum of 108 MHz to 684 MHz dynamically across both the upstream and downstream based on traffic demands. Combining that flexibility with additional machine learning tools to anticipate network issues could go a long way to restoring customers’ faith in the Rogers network.

It’s also worth noting that Comcast’s licensing arrangement also provides for the management of fiber networks using both headend OLTs and remote OLT modules. So, Rogers and any other potential licensees could adopt the framework across both their DOCSIS and fiber footprints.

Will Comcast Technology Solutions License to Additional Cable Operators?

The ongoing (and now very real) threat of Comcast’s entry into the broadband access technology licensing game has certainly disrupted the vendor landscape. If you have not been supplying Comcast with vCMTS, RPD, amplifiers, or other technologies, then the TAM for your products certainly takes a hit, especially now with the combined entity of Rogers and Shaw being taken off the table.

The question now is this: Are there any other operators who could potentially license Comcast’s broadband access solution? The obvious candidates are the operators who have been licensing Comcast’s X1 video platform. Besides Rogers and Shaw, these include Cox Communications and Videotron.

At this time, however, we don’t believe any of these remaining operators are interested in licensing Comcast’s broadband architecture and services. X1 was timely because it provided an advanced UI and backend video management platform for a service (broadcast TV) that was hemorrhaging subscribers amidst increasing content costs. Broadband is not in that same situation. Plus, operators are far more reluctant to potentially cede roadmap control to another operator—especially since they have already been doing that in some cases with their equipment vendors.

But if operators continue to have difficulty adding broadband subscribers, especially with competition increasing and margins potentially decreasing, then that could open the door for Comcast to expand its broadband access licensing footprint. Just as Broadcom has made its unified DOCSIS 4.0 chipset available to all operators in an effort to build scale, Comcast is looking to build a similar scale for its offering. It won’t be a significant money-maker for the operator, but more a mindset and market-maker.

2024 is shaping to be a year of correction across most enterprise network markets, as indicated by a notable decline in spending among key segments. This correction follows extraordinary growth from 2021 to 2023, driven by the surge in demand caused by the pandemic. Now, as enterprises work through the backlog of pandemic-driven investments and face excess inventory, coupled with cautious spending due to ongoing economic uncertainty, the five key enterprise network markets – Network Security, Branch Routing, Campus Switching, WLAN, and Enterprise Data Center Switching – are poised for varying degrees of growth deceleration and all but one an outright contraction.

Common Trends Across Enterprise Networking Markets

Across all five segments, a significant driver of the 2024 correction is a period of “enterprise digestion,” where organizations deploy the substantial purchases made during the pandemic and subsequent supply chain recovery periods. This digestion phase is compounded by excess inventory in the channel, leading to a slowdown in new equipment deliveries.

Moreover, macroeconomic factors such as inflation and tightening IT budgets are dampening network infrastructure investments. The broader technology market faces some pullbacks as companies reevaluate spending in light of potential economic uncertainties. For many vendors, the tailwinds provided by large backlogs of orders during the pandemic have dissipated, and there is now a recalibration as the market waits for demand to return to a more normalized, pre-pandemic level.

Network Security: Adjusting After High Growth

The Network Security market, with firewalls as its largest segment, has faced significant headwinds in 2024 as enterprises work through existing hardware investments. The slowdown reflects a broader trend in the industry, where organizations, having invested heavily in physical firewall solutions during the pandemic, are now focused on optimizing those assets rather than purchasing new hardware. At the same time, growth in non-hardware solutions like Security Service Edge (SSE) and virtual firewalls has helped cushion the decline, but even SSE is seeing a deceleration. After years of explosive growth, SSE investments are beginning to normalize as enterprises slow their spending to integrate existing deployments fully. This shift signals a cooling from the rapid pace of adoption seen in prior years, though the demand for flexible, cloud-based security solutions remains vital for the long term.

Branch Routing (SD-WAN and Access Routers): Temporary Slowdown Amid Strategic Shifts

The Branch Routing market, which encompasses SD-WAN and traditional access routers, is experiencing a slowdown in 2024 as enterprises take a strategic pause following the rapid expansion of these technologies during the pandemic. SD-WAN saw significant growth as organizations took the opportunity to invest in branch transformation to provide a better network experience at lower TCO at the branch. Still, this surge has now led to inventory overhangs. Additionally, the ongoing integration of SD-WAN functionality into broader Secure Access Service Edge (SASE) frameworks has shifted purchasing behavior. Enterprises are focusing on consolidating and optimizing their existing deployments rather than making new investments, contributing to the temporary softness in the market.

Campus LAN (WLAN and Campus Switching): Post-Pandemic Normalization

The Campus Switching and WLAN markets are enduring a similar correction as the post-pandemic glut of equipment deliveries is digested. After enjoying strong growth from 2021 to 2023, WLAN sales have contracted in 2024 as enterprises and distributors have worked through high stock levels accumulated during the supply chain recovery.

For Campus Switching, the slowdown has also been dramatic, especially in North America, where revenues have been dropping sharply following a peak in 2023. Excessive backlogs cleared in 2023 have led to a steep decline in new orders. Still, the rise of Power over Ethernet (PoE) technology and higher-speed ports, such as 2.5 Gbps and 5 Gbps, could offer growth opportunities as organizations prepare their networks for Wi-Fi 7, whose adoption began to accelerate in 2Q24.

Enterprise Data Center Switching: The Weakest Segment

The Enterprise Data Center Switching market has been hit the hardest in 2024 thus far. Despite some growth in the broader data center market due to AI-related investments, the enterprise segment has struggled as traditional front-end deployments face intense inventory challenges. The contraction is driven by prolonged backlog normalization and fewer large-scale deployments in non-cloud enterprise environments. The long upgrade cycles for enterprise data centers and a strategic pivot towards cloud and AI back-end networks have made this sector particularly vulnerable.

Outlook for 2025: A Return to Growth

Despite the contraction in 2024, the outlook for 2025 is brighter. The fundamental demand for digital transformation, cloud migration, and hybrid work solutions remains intact. As enterprises complete the digestion of their current investments and inventories normalize, spending is expected to rebound. The rise in adoption of AI-driven workloads, 5G, Wi-Fi 7, and advanced security frameworks like SASE will drive growth across the network infrastructure landscape. Furthermore, as inflation and interest rates decrease, enterprises will benefit from improved capital availability, providing further tailwinds for market growth.

However, the Enterprise Data Center Switch market is an exception, as it is not expected to return to growth in 2025. This segment is expected to face challenges. As enterprises continue to embrace the cloud or AI, they will increasingly be on public clouds, reducing the need to expand or refresh their on-premises data center footprint. As a result, further contraction is anticipated for enterprise data center switching, even as other markets recover and expand.

In conclusion, while 2024 is a year of necessary correction following unprecedented growth, the long-term prospects for enterprise network markets remain buoyant. Much of the industry is poised for recovery and growth in 2025, fueled by security, connectivity, and digital infrastructure innovations.

Two startups with shared advisory board members are hitting the market with solutions designed to facilitate cable’s convergence with wireless and 5G. The announcements are well-timed, given the funk cable operators find themselves in as net broadband subscriber losses mount, but also as those same operators continue to take a sizable share of new mobile subscribers. Additionally, the cable industry, which has for years benefited from decades of shared development cycles and deployments of the latest DOCSIS technologies, finds itself with multiple paths forward (DOCSIS 3.1+, DOCSIS 4.0, fiber, FWA, etc.) and trepidation that the next technology decision will leave it further behind its competitors.

Air5 and Air Wireless each aim to solve different problems cable operators face today. However, both share a core belief: cable operators’ future success depends on their ability to get to market quickly and build networks that transparently handle both fixed broadband and wireless traffic and services across their networks. In a recent blog, I detailed how US telcos are betting on service convergence to continue to chip away at cable’s massive broadband subscriber base. It stands to reason that cable operators will fight back using the same approach.

Extending DOCSIS Wirelessly

First, Air Wireless is pitching a solution that allows cable operators to extend their DOCSIS networks and services wirelessly using E-band spectrum, ranging from 60 GHz to 90GHz, and a point-to-multipoint architecture that looks and feels very similar to how optical nodes are distributed throughout an HFC network. The technology isn’t new. In fact, Air Wireless acquired the assets from a Slovenian startup known as Globtel, which had developed the Gigaray platform to transport voice, video, and DOCSIS data traffic wirelessly from a base station to transceivers located at businesses, MDUs, and residences. The transceivers connect to existing DOCSIS 3.1 modems and set-top boxes, allowing for a quick and easy method for aggregating and backhauling DOCSIS traffic.

The primary benefit of the Air Wireless solution to operators is time-to-market. Operators can extend their DOCSIS networks without having to run fiber to a new node location. Or, an operator can deploy the solution as a way to get services to an MDU or new neighborhood quickly and in advance of a more traditional buildout of an HFC network. In rural areas or regions where the costs associated with deploying fixed infrastructure just don’t make sense relative to subscriber ARPU, the Air Wireless solution gives operators a more cost-effective option for DOCSIS network extensions. Because of this flexibility, the company is reported in customer trials around the globe.

In the US, the key opportunity lies in the upcoming BEAD-, RDOF-, and Capital Projects Fund-related rollouts, which are time-sensitive and aimed at addressing lower-density rural and underserved areas. In India, cable operators such as Hathway, Den, and others are seeking ways to expand their networks and remain competitive with Reliance Jio and Bharti, both of which have begun significant fiber expansions. The Indian government continues to subsidize rural broadband rollouts to remote villages, where the Air Wireless solution could play a role in distributing broadband services. In Europe, where permitting delays and labor costs make network expansions costly, the Air Wireless solution could be used to extend DOCSIS networks more quickly.

One of the more interesting applications for the Air Wireless solution that also has global appeal is using the platform as a way to overbuild and upgrade existing HFC plants to deliver end-to-end DOCSIS 3.1 capabilities and take advantage of the more flexible modulation formats offered by OFDM. Many operators are still using DOCSIS 2.0 and DOCSIS 3.0, in some cases without channel bonding. Instead of potentially swapping out amplifiers or doing faceplate upgrades for new diplex filters, operators could use the Air Wireless platform with Remote PHY or Remote MACPHY modules to move to DOCSIS 3.1 more cost-effectively. In Latin America, for example, where cable operators are moving to fiber instead of upgrading from DOCSIS 2.0 or 3.0 to DOCSIS 3.1, the Air Wireless platform could give them a more cost-effective way to add throughput without the significant labor costs associated with trenching fiber.

Converging DOCSIS and 5G

While Air Wireless is focused on extending DOCSIS networks wirelessly, Air5 is focused on converging DOCSIS and wireless networks, taking advantage of architectural similarities between mobile backhaul networks and DAA-based DOCSIS networks. The CU (Centralized Unit) and Distributed Unit (DU) of 5G networks are roughly equivalent to the Remote PHY, Remote MACPHY, and select functions of the vCMTS in DAA networks.

Ultimately, the vision is that optical nodes become small cell sites with a shared infrastructure allowing cable operators to continue delivering DOCSIS data services as they do while also either continuing to offload their MVNO mobile traffic onto their Wi-Fi networks or directly onto the converged network via radio units that can handle the frequency conversion required to hand off mobile traffic. The shared infrastructure will require an upgrade to existing outside plant equipment so that DOCSIS data can still be delivered in spectrum up to 1.2 GHz, while 5G traffic can be transported anywhere between 3 GHz-5 GHz. New amplifiers, which Air5 is working on with partners, will have to be deployed. That might be a hard pill to swallow for operators who are just about to upgrade much of their installed amplifier base to 1.8 GHz.

Fixed-mobile convergence has been in various stages of discussion and deployment for years if not decades. So, why is this time different? Let’s consider a few different reasons:

Mobile subscriber growth and service bundling are critical for cable operators. In the US, the largest cable operators have seen significant growth in their mobile subscriber numbers, providing a silver lining to the dark cloud of broadband subscriber losses. Cable operators have grown their mobile subscriber base via MVNO relationships with Verizon and T-Mobile, but they are increasingly looking to deploy their own CBRS spectrum to become more self-reliant. Service bundling—especially if it allows subscribers to do truly seamless hand-offs between 5G and Wi-Fi networks while maintaining a single subscriber identity—is a critical goal of all operators.
Cable operators have powered outside plants. One of the biggest arguments against HFC networks, when compared with PON-based fiber networks, is actually a significant advantage when it comes to convergence: Power. HFC networks rely on signals that need to be amplified approximately every 2500 feet. To support this, 90-volt AC power inserters have been deployed at consistent intervals to provide for the powering of nodes, amplifiers, and Wi-Fi access points. In fact, US cable operators have deployed over 600 K Wi-Fi access points partially due to the availability of power at strategic locations. Cable operators not only have enough power to deploy small cells but also the fiber necessary to backhaul these small cell sites.
Control and user plane separation makes convergence easier. Because 5G core networks provide control and user plane separation, it becomes easier to converge 5G and Wi-Fi networks across the RAN and core. Additionally, cable operators’ transition to DAA architectures helps to virtualize DOCSIS networks. This gives operators much greater flexibility to offer network slicing, allowing Wi-Fi traffic can ultimately be managed by a converged 5G and DOCSIS core. This process begins with an evolution of the vCMTS to a vBNG and then an AGF (Access Gateway Function), which essentially serves as the bridge between the wireline network and the mobile core.

Expanding the Component Vendor Ecosystem

One of the benefits of convergence is the potential increase in the number of component vendors developing new chips to support the larger, combined TAM (Total Addressable Market.) There is probably no segment in the communications sector that could use a supplier expansion other than DOCSIS, which has historically been dominated by Broadcom. In fact, in a recent blog, we argued that Broadcom’s decision to accelerate the availability of a 3 GHz-capable unified chip that supports DOCSIS 5.0 could be an effort to “pre-empt efforts by upstarts such as Air5, which is developing products that fuse 5G and DOCSIS networks and, simultaneously, opening up the shrinking DOCSIS component ecosystem to suppliers in the RAN and mobility sectors.”

We have already seen significant consolidation of DOCSIS infrastructure and CPE suppliers in the last year and we fully expect that this will continue, as the DOCSIS equipment TAM, by itself, is not enough to sustain the current vendor ecosystem. Component supplier consolidation is expected soon, as well, certainly with Qualcomm’s rumored exploration of an acquisition of Intel.

Lurking around are the likes of Nvidia and AMD, who are looking to merge signal processing and GPUs. Though these components would be designed for use in mobility networks, there is no reason they couldn’t be adapted to work in converged networks, as well.

Contact Us