TBR Spotlight Reports represent an excerpt of TBR’s full subscription research. Full reports and the complete data sets that underpin benchmarks, market forecasts and ecosystem reports are available as part of TBR’s subscription service. Click here to receive all new Spotlight Reports in your inbox.

AI data needs raise expectations about how services vendors and hyperscalers should be thinking about their relationships, with region-specific guardrails further testing joint GTM durability

Key trends
Hyperscalers’ dispersion across the globe requires these vendors, along with their services partners, to carve out new operating models that prioritize both go-to-market success and adherence to local requirements. This balance translates into a case-by-case — or rather, country-by-country — approach with local partnerships being the common thread. This is especially true for hyperscalers competing for local market share — particularly in the European Union (EU) — to the extent of relinquishing operational control (e.g., Microsoft Bleu), creating a direct opportunity for regional system integrators (SIs) and local infrastructure operators. While global SIs (GSIs) are also in contention, they must account for these vendors as they build out local presence, along with their commercial, staffing and partner models. Meanwhile, regional legislation is pushing hyperscalers to commit to investment pledges to ensure business continuity, with implications for services partners and buyers, further testing the limits of their relationships.

Go-to-market strategy
Enterprise buyers are becoming increasingly conflicted in their expectations of how vendors can best support their technology needs. When it comes to interoperability, many customers look to leverage third-party vendors, and a smaller percentage expect their cloud vendors to address these concerns directly. At the same time, buyers continue to identify technology expertise as a key skills gap in vendors’ value proposition across regions. These dynamics are further amplified in vendors’ regional go-to-market strategies, especially when it comes to accounting for the role of AI and niche vendors that bring specialized knowledge. In a nutshell, vendors cannot rely on a one-size-fits-all AI ecosystem strategy across regions. Success will require region-specific approaches: IP-led initiatives in APAC, orchestration frameworks in Europe, and startup-centric marketplaces in the Americas. All must be underpinned by interoperable APIs and strong governance to help IT services providers capture and monetize local demand. Executing against such expectations while continuing to rely on a traditional labor-arbitrage model will test professional services firms’ readiness to transform their own operations while maintaining trust with hyperscalers, which continue to explore the opportunity to drive professional services revenue by simplifying the sales process and marketplace through the use of agentic AI.

Vendors
In addition to launching bespoke operating models and investing in local infrastructure, hyperscalers are investing at the infrastructure layer to support workload portability and management capabilities that help customers adhere to sovereignty regulations. Google Distributed Cloud is probably the most sovereignty-forward example, but similar comparisons can be made for both Microsoft and Amazon Web Services (AWS). Security is another critical area hyperscalers are investing in, whether it is Microsoft’s new deputy chief information security officer (CISO) role for Europe or Google’s continued effort to expand Mandiant’s assets to prevent breaches. Orchestrating these evolving offerings into a cohesive IT estate will be an opportunity for services vendors, especially when deciding how to configure them in a way that is suitable for country-specific needs. Services — including migration, implementation, consulting and advising — all play a role in navigating the increasingly complex regulatory and security environment, requiring broad hiring in the EU, where the opportunities are the greatest.

Data location will remain a leading barrier to cloud adoption, but interoperability and breaking down data barriers across platforms will present the bigger opportunity for services vendors

As highlighted in TBR’s 2H24 Cloud Infrastructure & Platforms Customer Research, data location ranked as the second-highest cloud pain point after security, with 40% of respondents expecting their cloud vendors to directly address these concerns — largely due to vendors doing a better job of making customers aware of the various data center hosting and encryption options. Conversely, when it comes to interoperability, many customers will leverage third-party vendors, and a smaller percentage expect their cloud vendors to address these concerns directly. This speaks to both the skills that services-led firms have amassed across multiple technology platforms and the high degree of lock-in the hyperscalers still create across their infrastructure. That said, from a technology perspective, vendors are doing a better job of integrating their offerings, particularly in the area of agentic AI, which necessitates more robust data sharing. As more open frameworks, including Google’s A2A (Agent2Agent), mature and become enterprise-ready, they could create a needed level of standardization that GSIs can leverage to build new agents alongside their ISV partners. Deloitte’s partnership with Google Cloud to build ServiceNow-specific agents on A2A is a good example.

“It [sovereign cloud] helps, significantly helps. A completely separated, air-gapped environment different from the regular public cloud itself. And Switzerland, by the way, is a great example. AWS put up two regions in Switzerland. It’s a tiny country. But even then, they had two regions to satisfy this condition of resilience, etc.” — Managing Director (Firmwide) & Chief Data Architect, Financial Services

Cloud vendor insights excerpt

Microsoft’s Cloud Services & Ecosystem Strategy

Microsoft Cloud’s Latest Ecosystem Initiative (Source: TBR 1H25)

Microsoft Cloud’s Estimated Ecosystem Statistics for 1H25 (Source: TBR)

TBR Spotlight Reports represent an excerpt of TBR’s full subscription research. Full reports and the complete data sets that underpin benchmarks, market forecasts and ecosystem reports are available as part of TBR’s subscription service. Click here to receive all new Spotlight Reports in your inbox.

CSPs have an opportunity to capture meaningful value from AI, but realizing this opportunity requires action and investment

Most CSPs look for quick-hit ROI from GenAI; there is still hesitancy to commit to larger-scope AI initiatives that require significant upfront investment

Leading CSPs are all dabbling with GenAI, with some use cases already commercially scaling, especially in customer care and BSS.

CFOs at some leading CSPs are getting directly involved in AI programs to give visibility to these initiatives and ensure they are paying off, looking for quick ROI.

So far, CSP managers like what they are seeing and are hopeful about the prospects of AI delivering significant outcomes, but this hope has yet to translate to large-scale investments tied to broad transformations. Rather, CSPs are focused on more tactical, smaller-scope solutions that address specific pain points or that promise fast ROI.

TBR believes it will take some time for most CSPs to evolve this investment behavior, but some other considerations also need to be factored into the equation, such as uncertainty about governance, regulation and data efficacy. For example, AI’s effectiveness correlates closely with the volume and quality of the data input into the model. The reality is that data inside CSP organizations are usually highly disaggregated in silos and on legacy systems, which poses a major challenge to undergoing large-scale AI transformations.

Realizing the $170 billion total opportunity TBR estimates AI presents the telecom industry by 2030 requires CSPs to act differently

AI presents a once-in-a-generation opportunity for the telecom industry to achieve two key objectives: generate new revenue and reduce costs.

However, there is real risk that most CSPs globally will miss out on the AI opportunity due to cultural (behavioral) and regulatory encumbrances, such as long decision cycles, an unwillingness to invest what it takes to win and general risk aversion. These encumbrances, which are endemic to the telecom industry, have resulted in CSPs largely missing out on every major opportunity in the last two decades (e.g., cloud computing opportunity, video streaming opportunity, digital advertising opportunity), many of which were won by the hyperscalers that were willing and able to take these risks.

Though leading CSPs have been investing in AI, TBR notes that most of these investments seem to be myopically focused on quick-hit wins, which is acceptable in the short term, but true opportunity capture will be contingent on broader-scope initiatives, coupled with upfront investment.

CSPs cannot afford to miss out on another opportunity, especially one that has such transformational qualities as AI. Getting AI right should be of paramount concern to CSPs, as competitors that transform with AI may obtain unapproachable differentiation compared to CSPs that do not invest in AI. Said differently, CSPs that do not get AI right may not be competitive anymore, leading to longer-term questions about their viability.

CFOs need to evangelize AI across their organizations to seize this critical opportunity. True digital transformation with AI will require significant upfront investment, and the payoff of some investments take a longer time to realize than others. CFOs need to balance investment levels with outcomes, but more support is needed from the top to approach AI in a broader manner.

TBR notes that of the $170 billion CSP AI opportunity by 2030, an estimated 50% is likely to be realized in APAC, where there are many CSPs aiming to take a leadership role in the AI market. The lack of domestic hyperscalers (outside of the U.S. and China) and government desire to have sovereignty over technologies of national importance gives CSPs an opportunity to fill in the gaps in the market. China-based CSPs are already diving headfirst into the AI opportunity, and they are likely to generate billions of dollars in new revenue and save billions of dollars in costs from AI by 2030.

Telcos have a data usability problem, which could slow the pace of AI adoption in their businesses; addressing this issue requires significant additional investment

CSPs have a data preparedness problem, which is an underestimated issue and will be costly to rectify

The efficacy of and value provided by AI are contingent on the quality, type and volume of data the AI model is trained on. It is therefore imperative that CSPs ensure their data is ready for AI and that the data is managed and governed in a sustainable way. This is especially pertinent now that GenAI can train on a broad range of structured and unstructured data.

The reality is that CSPs’ data does not reside in one data lake. Rather, it is highly disaggregated and siloed across the organization and, in some cases, may be incomplete. There are also restrictions (e.g., regulatory, privacy) and security considerations around certain types of data, which hinder the utility and accessibility of the data, creating another major challenge that companies will have to address before they can appropriately leverage AI.

Individual CSPs have access to petabytes of data, but most of this data resides in legacy systems and/or is subject to other restrictions, creating challenges around pooling and preparing the data to be leveraged for AI.

To rectify this issue, CSPs will need to clean up their technical debt and assemble their data in a single, unified platform. This horizontal data layer will be required to run and realize AI-driven outcomes at scale. Most CSPs will likely need assistance with this task, opening up opportunities for the vendor community.

Addressing this issue will also require significant investment and human resources, and TBR believes CSPs that are in the initial stages of their GenAI initiatives have underestimated or overlooked the problem.

TBR believes this issue could extend CSPs’ timelines for AI commercialization. Vendors and hyperscalers will try to mitigate the data issue by training AI models on their own datasets, which could also be fed by partners’ data, but there is a risk that this will not be sufficient and that AI models will need to be tuned to a specific CSP’s needs to realize desired outcomes. TBR also believes it will be common for stakeholders in the telecom ecosystem to be protectionist regarding their data, not wanting to (or being able to) share their data for fear of regulatory reprisals and a nullification of competitive advantages.

Sovereign cloud requirements have the potential to be a meaningful revenue driver for CSPs in select regions and countries

Countries (e.g., China, Saudi Arabia, Qatar, United Arab Emirates, South Korea) and alliance entities (e.g., NATO) that are more impacted by geopolitics and/or have strong data privacy rules and regulations (e.g., European Union [EU] member countries) are looking into establishing national AI infrastructure (e.g., data centers), also referred to as sovereign cloud.

Sovereign clouds would theoretically be set up and managed by domestic players for national security reasons and could be an area of opportunity for CSPs that reside in those countries or regions, whereby the CSP might host the cloud and provide some value-added services to manage those environments on behalf of the government and related entities.

However, TBR ultimately believes it is more likely that U.S. hyperscalers will participate in some way in the AI value chain for most countries given the broad scope of their involvement and dominance in the global AI ecosystem, despite intentions for national sovereignty. This includes CSPs partnering with hyperscalers to jointly develop, operate and manage cloud resources in-country, as evidenced by Orange’s use of Microsoft’s cloud and AI technology for its sovereign cloud joint venture with Capgemini (called Bleu) in France.

TBR estimates the potential annual AI-related opportunity for CSPs will reach $170B by 2030, approximately 53% of which is new revenue and 47% is cost efficiencies

CSPs have been largely sidelined from the new revenue opportunity presented by AI since the emergence of GenAI in 4Q22, but this is starting to change, evidenced by significant deals won by Lumen and Zayo to provide transport between data centers for AI workloads. There are also some green shoots of demand for hyperscalers to leverage CSPs’ network facilities (e.g., wire centers and central offices) and other real estate assets to colocate AI infrastructure closer to end users, as evidenced by efforts being made by Verizon and AT&T.

All CSPs that are investing in AI currently expect to reap cost efficiencies from the technology. The new revenue opportunity is more nuanced and is CSP- and market- specific in nature. APAC-based CSPs are likely to be the largest beneficiaries of new revenue from AI due to government protections, stimulus and cultural orientations toward early adoption of emerging technologies.

Total Annual Potential Value of AI to CSPs by 2030 (Source: TBR)

TBR Spotlight Reports represent an excerpt of TBR’s full subscription research. Full reports and the complete data sets that underpin benchmarks, market forecasts and ecosystem reports are available as part of TBR’s subscription service. Click here to receive all new Spotlight Reports in your inbox.

Interest in AI capabilities has not waned as enterprises view the technology as critical to long-term competitive positioning

The buzz around GenAI persists as enterprise interest is leading to adoption. Yet it is still early days, and many enterprises remain in exploration mode. Some use cases, such as data management, customer service, administrative tasks and software development, have already moved from the proof-of-concept stage to production. Still, the exploration phase of AI adoption will be a slow burn as enterprises seek opportunities beyond these low-hanging fruit. As seen in the graph to the right, most enterprises are evaluating AI qualitatively, forgoing quantitative measures to keep up with peers based upon the assumption that the technology will bring transformational improvement to business operations.

Source: TBR 2H24

Reasoning models excel at performing complex, deterministic tasks, and have become the most popular models at the back end of agentic AI

The capability improvement brought by the iterative inferencing process has made reasoning models the focal point of frontier model research. In fact, most of the models sitting atop established third-party benchmarks are reasoning models, except for OpenAI’s GPT-4.5, which the company stated would be its last nonreasoning LLM. Put simply, the difference in output quality is too pronounced to ignore, especially regarding complex, deterministic tasks. As seen in the graph, reasoning models outperform their nonreasoning predecessors across the board, with the greatest distinction appearing in coding and math benchmarks. The strength in complex, deterministic tasks makes reasoning models particularly adept at powering agentic AI capabilities, offering a wider range of addressable use cases and greater accuracy. In addition, reasoning frameworks can be leveraged at any parameter count, with available reasoning models ranging from fewer than 10 billion parameters to more than 100 billion.

As SaaS vendors continue to build proprietary, domain-specific SLMs [small language models] to power their agentic capabilities, incorporating reasoning frameworks will be an important part of their development strategies. Although the capabilities of reasoning models are impressive, the models bring new challenges and are not necessarily the best choice for every application.

Simple content generation and summarization, for instance, do not necessarily require iterative inferencing. Moreover, the greater compute intensity caused by repeated processing at the transformer layer will compound existing challenges to scaling AI adoption. Not only will these models be more expensive to run for the customer, but they will also exacerbate the persistent supply shortages facing cloud infrastructure providers. Microsoft has noted infrastructure constraints as a headwind to AI revenue growth in the past several quarters, and the emerging need for test-time compute adds to these infrastructure demands. As discussed in TBR’s special report, Sheer Scale of GTC 2025 Reaffirms NVIDIA’s Position at the Epicenter of the AI Revolution, NVIDIA’s CEO Jensen Huang stated that reasoning AI consumes 100 times more compute than nonreasoning AI. Of course, this was a highly self-serving statement, as NVIDIA is the leading provider of GPUs powering this compute, but we are dealing with magnitudes of difference. For the use of reasoning models to continue scaling, this high compute intensity will need to be addressed.

SaaS vendors will need to get on board with the new Model Context Protocol to ensure customers can use their model of choice

SaaS vendor strategy assessment

From a strategic positioning perspective, TBR does not expect the rising popularity of the Model Context Protocol to have an outsized impact, primarily because we anticipate all application vendors will adopt the framework to ensure customers can leverage the model of their choice. Furthermore, cloud application vendors are positioned to benefit from the standardization of API calls between models and their workloads. Through a standardized API calling framework, these vendors will be better positioned to drive cost optimization and improve workload management for embedded AI tools.

Recent developments

The Model Context Protocol is becoming the standard: The idea of the Model Context Protocol (MCP) has been steadily gaining popularity following its release by Anthropic in November 2024. At its core, MCP aims to address the emerging challenge of building dedicated API connectors between LLMs and applications by introducing an abstraction layer that standardizes API integrations. This abstraction layer — commonly referred to as the MCP server — would establish a default method for LLM function calling, which software providers would need to incorporate into their applications to access LLMs.

This standardization offers several benefits for model vendors, such as eliminating the need to build individual connectors for each service and promoting a modular approach to AI service integration, potentially unlocking long-term advantages in areas such as workload management and cost optimization.

For SaaS vendors, there is little reason to resist the shift toward MCP, and its growing popularity may make adoption inevitable. Application vendors like Microsoft and ServiceNow have already begun implementing the protocol by establishing MCP servers for the Copilot suite and Now Assist, respectively, and TBR expects other vendors to follow.

It is important to recognize, however, that this approach better suits vendors taking a model-agnostic stance — meaning they aim to empower enterprises to use any LLM to automate agentic capabilities. A possible exception lies with vendors that are less model-agnostic. For instance, Salesforce’s emphasis on proprietary models reduces the need for MCP and favors the company’s focus on native connectors between Customer 360 workflows and xGen models.

Ultimately, TBR expects Salesforce to adopt MCP, but there is an important distinction in how different SaaS vendors may approach standardization. Today, the BYOM [bring your own model] philosophy remains a priority for Salesforce, but if the company were to eventually push customers to use its proprietary models exclusively with Customer 360, its commitment to MCP could be deprioritized in favor of tighter customer lock-in.

Google enhances AI capabilities with the launch of Gemini 2.5 Pro, revolutionizing search functionality, healthcare solutions and multimodal content generation

Google remains differentiated in the AI landscape through the deep integration of its proprietary models across a broad product ecosystem, including Search, YouTube, Android and Workspace. Although many competitors focus on niche capabilities or open-source development, Google positions Gemini as a comprehensive, multimodal foundation model designed for widescale consumer and enterprise adoption. Google’s infrastructure, proprietary TPUs (Tensor Processing Units), and access to vast and diverse data sources provide a significant advantage in training and deploying next-generation models. Gemini 2.5 Pro is a testament to this strength, offering the best performance and largest context window available on the market. Although TBR expects the top spot to continue exchanging hands, we believe Google’s models will remain among the frontier leaders for years to come.

OpenAI advances AI development with GPT-4.5, cutting-edge agent tools and a premium ChatGPT Pro subscription to expand capabilities and improve user experiences

OpenAI is the most valuable model developer in the market today, largely due to the company’s success in productizing its models via ChatGPT. The mindshare generated by ChatGPT is benefiting the company’s ability to reach custom enterprise workloads, though OpenAI must be mindful of the widening gap in price to performance relative to peers. From a sheer performance perspective, TBR believes the company’s emphasis on securing compute infrastructure via the Stargate Project, as well as its ongoing partner initiatives to gain access to high-quality training data, will ensure its models remain near the top of established third-party benchmarks over the long term.

TBR Spotlight Reports represent an excerpt of TBR’s full subscription research. Full reports and the complete data sets that underpin benchmarks, market forecasts and ecosystem reports are available as part of TBR’s subscription service. Click here to receive all new Spotlight Reports in your inbox.

Most operators will sustain wireless service revenue and connection growth in 2025 but face headwinds from macroeconomic challenges and Trump administration immigration policies

Most benchmarked operators sustained service revenue growth in 4Q24, driven by connection growth and higher ARPA

Total wireless revenue from benchmarked U.S. operators increased 4.5% year-to-year to $78.9 billion in 4Q24, mainly due to continued postpaid phone subscriber growth and higher average revenue per account (ARPA). Although the market is maturing, operators are maintaining postpaid phone net additions due to factors including population growth and more businesses purchasing mobile devices for employees. Higher ARPA is being driven by operators increasing connections per account, including from growing fixed wireless access (FWA) adoption, uptake of premium unlimited data plans and recent rate increases implemented over the past year.
 
Though most U.S. operators expect to continue to grow wireless service revenue and connections in 2025, they will face headwinds from factors including macroeconomic pressures (including layoffs within the private and public sectors and uncertainty around tariff impacts) and immigration policies under the Trump administration (including mass deportations).

U.S. operators increase focus on cross-selling mobile and broadband services

U.S. operators are focused on advancing their convergence strategies by offering plans bundling mobile and broadband services. The bundles create a stickier ecosystem to reduce churn long-term via the convenience of enrolling in broadband and mobility services from the same provider as well as by providing discounted pricing compared to purchasing those services separately.
 
Operators including AT&T, Charter, Comcast, T-Mobile and Verizon are growing their ability to offer these bundles via the expanding service availability of their broadband services (including wireline and FWA offerings). Operators are also targeting acquisitions to strengthen their convergence strategies, such as Verizon’s pending purchase of Frontier Communications and T-Mobile’s proposed joint ventures to acquire Metronet and Lumos. Cable operators also have significant opportunity to increase sales of converged services as a relatively low portion of cable broadband customers are enrolled in their service provider’s mobile offering.

AI is providing cost savings and revenue generation opportunities for U.S. operators

U.S. operators are focused on more deeply implementing AI technologies in areas including optimizing customer service and sales & marketing functions as well as enhancing network operations. For instance, deeper AI implementation will help AT&T reach its goal of generating $3 billion in run-rate cost savings between 2025 and 2027, while leveraging AI technologies will help T-Mobile meet its target of reducing the number of inbound customer care calls by 75%.
 
AI will also help operators optimize energy usage, especially as it pertains to network operations. Examples include using AI for optimal, dynamic traffic routing and to determine when to turn on and turn off radios to optimize energy usage. AI, especially providing network and real estate resources to support AI inferencing workloads, will create significant revenue opportunities for operators.
 
For instance, Verizon views telco AI delivery as having a $40 billion total addressable market, and the company has already secured a sales funnel of over $1 billion in business by leveraging its existing infrastructure and resources.

Operators are focused on cost-cutting initiatives, including streamlining headcount and more deeply implementing AI technologies, to improve margins

The impacts of inflation and challenging macroeconomic conditions, such as lower consumer discretionary spending, higher network operations and transportation expenses, and increased labor-related costs, are limiting profitability for U.S. operators. These challenges are leading operators to implement cost-cutting and restructuring initiatives to improve profitability, such as AT&T’s goal of generating $3 billion in savings from 2025 to the end of 2027 through its latest cost-cutting program.
 
Operators are streamlining headcount as part of their cost-cutting initiatives. For instance, about 4,800 employees are expected to leave Verizon by the end of March 2025 as part of the company’s latest voluntary separation program.
 
To increase cost savings and operational efficiencies, operators are more deeply implementing AI technologies in areas including customer service, field technician support and fleet vehicle fuel consumption.
 
T-Mobile is improving profitability, evidenced by its EBITDA margin growing by 220 basis points year-to-year to 35.4% in 4Q24, which was impacted by the company’s higher revenue and lower network costs aided by greater merger-related synergies. T-Mobile’s 2025 guidance for core adjusted EBITDA* is between $33.1 billion and $33.6 billion, compared to $31.8 billion in 2024. Service revenue growth as well as cost-cutting initiatives and merger-related synergies will all contribute to higher core adjusted EBITDA.
 
*Core adjusted EBITDA reflects T-Mobile’s adjusted EBITDA less device lease revenues.
 

Wireless Revenue, EBITDA Margin and Year-to-year Growth for 4Q24 (Source: TBR)

T-Mobile continued to lead the U.S. in postpaid phone and broadband net additions in 4Q24 and recently launched new FWA pricing plans

Operators are attracting FWA customers, mainly because FWA offerings have lower price points compared to other broadband services and are available to customers in markets with limited other high-speed broadband options, such as within rural markets. Though consumers account for the bulk of FWA connections, FWA is also gaining momentum among businesses seeking to reduce connectivity expenses and/or companies needing to quickly launch new branch locations, as the technology can be installed faster than fixed broadband.
 
In 4Q24 T-Mobile continued to lead the U.S. in broadband subscriber growth, driven by its FWA services, aided by the company continuing to gain market share against cable companies including Comcast and Charter, which reported steeper broadband customer losses in 4Q24 both year-to-year and sequentially. T-Mobile also reported its highest-ever year-to-year broadband ARPU growth in 4Q24, which was aided by the company revamping its 5G Home Internet and Small Business Internet plans in December.
 

Total FWA Net Additions for 4Q23-4Q24 (Source: TBR)

Wireless capex moderated for most U.S. CSPs in 2024 as they are in the later stages of 5G rollouts

Verizon’s consolidated capex will increase to a guidance range of $17.5 billion to $18.5 billion in 2025, compared to $17.1 billion in 2024 (higher consolidated capex is mainly due to increased wireline capex to support Verizon’s accelerated Fios build). TBR estimates Verizon’s wireless capex in 2025 will be relatively consistent compared to 2024 as the company will focus on the continued expansion of C-Band 5G services into suburban and rural markets.
 
AT&T’s 2025 guidance for capital investment, which includes capex and cash paid for vendor financing, is in the $22 billion range, consistent with $22.1 billion in capital investment in 2024. Capital investment in 2025 will entail materially lower vendor financing payments compared to 2024, while capex is expected to increase year-to-year in 2025. TBR estimates AT&T’s wireless capex will be about $10.6 billion in 2025, which will help to meet AT&T’s goals, including providing midband 5G coverage to over 300 million POPs by the end of 2026 and completing the majority of its transition to open-RAN-compliant technologies by 2027.
 
T-Mobile’s capex guidance for 2025 is around $9.5 billion, compared to $8.8 billion in capex spent in 2024, with spending focused on continued 5G network deployments as well as investments in IT platforms to enhance efficiency and customer experience.

TBR Spotlight Reports represent an excerpt of TBR’s full subscription research. Full reports and the complete data sets that underpin benchmarks, market forecasts and ecosystem reports are available as part of TBR’s subscription service. Click here to receive all new Spotlight Reports in your inbox.

Behind healthy server backlog and new software IP, hardware vendors drive the cloud components market, particularly as software pure plays prioritize entirely public cloud migrations

Hardware-centric vendors continue to make their move into software

Over the past several years, the cloud software components market has shifted. Microsoft and Oracle are no longer dominating the market as they prioritize their native tool sets and encourage customers to migrate to public cloud infrastructure. Driven largely by weaker-than-expected purchasing around Microsoft Windows Server 2025, aggregate revenue growth for these two software-centric vendors was down 3% year-to-year in 3Q24. Over the same compare, total software components revenue for the benchmarked vendors was up 14% and total cloud components revenue was up 8%. In some ways, this dynamic has made room for hardware-centric vendors such as Cisco and Hewlett Packard Enterprise (HPE) to move deeper into the software space, particularly as they buy IP associated with better managing orchestration infrastructure in a private and/or hybrid environment.

Backlog-to-revenue conversion for AI servers fuels market growth

Though revenue mixes are increasingly shifting in favor of software, driven in part by acquisitions (e.g., Cisco’s purchase of Splunk), hardware continues to dominate the market, accounting for 80% of benchmarked vendor revenue in 3Q24. Industry standard servers being sold to cloud and GPU “as a Service” providers are overwhelmingly fueling market growth, more than offsetting unfavorable cyclical demand weakness in the storage and networking markets. This growth is largely driven by the translation of backlog into revenue, but vendors are still bringing new orders into the pipeline, which speaks to ample demand from both AI model builders and cloud providers. However, large enterprises are increasingly adopting AI infrastructure as part of a private cloud environment to control costs and make use of their existing data.

Cloud Revenues by Segment for 3Q23-3Q24 (Source: TBR)

Ample scale and strong demand from both CSPs and enterprises extend Dell’s lead in the cloud components market

Cloud components vendor spotlights

Dell Technologies [Dell]

From a revenue perspective, HPE and Cisco once threatened Dell’s cloud components leadership, but the company has been able to distance itself from its nearest competitors. This is largely due to Dell’s performance over the past year, with strong server demand, particularly from Tier 2 cloud service providers (CSPs), propelling the company’s corporate and cloud components revenue growth rate to the double digits. Meanwhile, in 3Q24 Dell shipped $2.9 billion worth of AI servers while backlog reached $4.5 billion, reflecting 181% year-to-year growth during the quarter and indicating strong future revenue performance.

Hewlett Packard Enterprise

Like its peers, HPE is benefiting from AI-related server demand, and in 3Q24 the company reported $1.5 billion in total AI systems revenue. HPE continues to benefit from its ongoing efforts to shift the sales mix in favor of software and services via GreenLake. In 3Q24 HPE completed its acquisition of Morpheus Data, officially equipping HPE with a suite of infrastructure software that allows customers to take core hypervisors, such as KVM and VMware, and use them to build complete private cloud stacks.

Cisco

With its acquisition of Splunk, Cisco has emerged as the leader of the software components market, even surpassing Microsoft in related revenue. But networking still accounts for the bulk of Cisco’s components business, and, as evidenced by a 32% year-to-year decline in total hardware revenue for 3Q24, Cisco is facing headwinds in the core networking business. That said, the company is actively taking steps to build out its portfolio, particularly by integrating more security components into the networking layer, which is where most cyberattacks originate, to boost its long-term competitiveness in the market.

Infrastructure agnosticism and flexible cloud-enabled delivery are core attributes of the service delivery market, cementing IBM’s leadership

Dedicated orchestration tools continue to have their place in the market, both in on-premises and cloud environments, but growth is largely driven by application lifecycle management and orchestration tools that span multiple environments. IBM has a rich history in this space and remains a revenue leader. Cisco used to have a foothold in the market but no longer sells its CloudCenter suite.

Vendor spotlight: IBM

After taking steps to bring watsonx into Maximo in 2Q24 for greater process automation, IBM strengthened its commitment to the asset performance management space with the acquisition of Prescinto. Prescinto offers AI tools and accelerators designed for asset owners and operators with a focus on renewable energy and operators. This deal is designed to support IBM’s play in certain verticals, particularly energy and utilities.

Service Delivery and Orchestration Revenue Growth vs Cloud Software Components Revenue Growth for 3Q24 (Source: TBR)