TBR’s Consulting & Systems Integration Market Forecast provides a comparison across tracked vendors, establishes best practices and identifies what will separate leaders from laggards over the next five years.
TBR’s IT Services Market Forecast is a comprehensive, forward-looking analysis of the IT services market, including market share of leading IT services vendors and the competitive and customer dynamics driving revenue growth.
It is increasingly clear that the AI infrastructure market is neither unified nor single-dimensional. What began as a relatively cohesive, GPU-driven infrastructure build-out is rapidly diverging along three structurally different yet interrelated axes:
Together, these dynamics represent fundamental shifts in how AI servers and systems are architected, deployed and monetized, impacting the entire AI ecosystem.
As these divides deepen, the AI infrastructure market is effectively splitting into distinct segments with different buyers, economics and competitive dynamics. Vendors and customers that continue to treat AI infrastructure as a single, homogeneous market risk misallocating capital, overestimating growth opportunities and underestimating emerging competitive threats.
Perhaps the most important divide in the AI infrastructure market is between training and inference workloads. Training prioritizes flexibility, scalability and rapid iteration, reinforcing GPUs as the foundation for frontier model development. Inference, however, operates under different constraints, where cost efficiency, power consumption and throughput take precedence, especially for hyperscalers. Flexibility remains important across more variable enterprise and heterogeneous workloads. This inference dynamic is driving hyperscalers to increasingly deploy custom AI ASICs optimized for cost per inference and energy efficiency.
TBR sees this shift as a reflection of a broader change in where value is created rather than an indication of a transition from one architecture to another. While training has driven initial infrastructure build-outs, inference represents the larger long-term opportunity as AI adoption expands across industries. As a result, the center of gravity in AI infrastructure is shifting toward production inference workloads, where efficiency and scale define competitiveness at the hyperscaler level, even as flexibility remains a key requirement across the broader market.
TBR AI Server & Systems Market Forecast by Accelerator Type (Source: TBR Estimates)
The divergence between GPUs and custom AI ASICs reflects how different ecosystem players are positioning to capture this growing inference opportunity. Hyperscalers are investing in custom silicon to optimize performance, reduce costs and align infrastructure with large-scale, stable workloads, while also abstracting infrastructure through managed services to consolidate control higher up the stack.
At the same time, merchant accelerator vendors, like Advanced Micro Devices (AMD) and NVIDIA, are not ceding the inference opportunity to hyperscalers and their custom AI ASICs. Instead, they are investing aggressively to bolster their platform-level capabilities through tightly integrated hardware and software stacks, managed infrastructure offerings and next-generation systems optimized for tokens-per-watt efficiency. Competition shifts from a hardware-centric model to a platform-level battle, where control over how AI infrastructure is delivered, consumed and monetized increasingly determines value capture and directly influences deployment models.
While hyperscalers remain the largest demand vector behind the growing AI infrastructure market, the idea that AI workloads will be fully centralized in the cloud is already beginning to break down. Enterprises are encountering practical constraints, including data sovereignty requirements, data gravity, latency sensitivity and cost considerations, that are driving a more distributed deployment model.
At the same time, many organizations continue to face challenges related to data readiness and infrastructure complexity, which slow large-scale enterprise AI adoption and reinforce the need for hybrid approaches. As a result, TBR sees organizations deploying AI infrastructure across a mix of cloud, on-premises and edge environments — with that mix dictated by industry group and company size — rather than converging on a single deployment model.
This dynamic echoes previous technology cycles, in which organizations choose hybrid cloud architectures rather than centralizing all workloads in public or private cloud environments. However, TBR believes AI will create an even more fragmented and distributed infrastructure landscape, where deployment decisions are more closely tied to workload-specific requirements.
As these structural divides take hold, vendors are already being forced to make strategic trade-offs.
OEM strategies, for example, are diverging between high-volume, lower-margin deals with services providers and more targeted, higher-margin enterprise opportunities that emphasize integrated solutions and services. TBR views this divergence as a reflection of the broader realities that there is no single, unified go-to-market strategy for AI infrastructure and that demand and adoption by customer group is uneven. As such, vendors must align their current portfolios and alliance and investment strategies with specific market segments to optimize value capture rather than attempting to compete across all fronts simultaneously.
Upstream of the OEMs, NVIDIA’s near-monopoly position in the AI infrastructure market is gradually receding as hyperscaler AI ASICs and other merchant accelerators vie for their place in the AI data center. AMD’s investments in developing rack-scale integrated systems and emphasis on ecosystem openness directly compete with NVIDIA in the merchant accelerator space, while hyperscaler AI ASICs pose an adjacent threat for share of the inference market. Peripherally, other vendors are also entering the merchant market with processors and systems architectures purpose-built for specific inference applications.
As fragmentation accelerates, competitive positioning will increasingly depend on market segment alignment.
In this environment, there is no single winner across AI infrastructure. Instead, leadership will be defined within each segment, and success will depend on how effectively vendors align their strategies with the underlying structure of the rapidly evolving market.
Understanding how AI infrastructure is fragmenting — and where value is shifting as a result — is critical for forecasting demand, evaluating competitive positioning and aligning long-term strategy.
TBR’s AI Infrastructure Market Landscape provides a detailed analysis of these dynamics, including vendor performance, ecosystem developments and evolving market opportunities across the AI infrastructure stack. Preview the data and analysis in our latest AI Infrastructure Market Landscape.
All hyperscalers tout themselves as “full-stack” to a degree, but Google Cloud’s distinct advantage is that it owns a leading frontier model, Gemini. Having Gemini deeply embedded throughout the portfolio creates a powerful flywheel effect that lets Google Cloud monetize AI in ways others cannot. At the same time, the value is shifting from the AI models themselves to how those models work with a growing set of tools and data to create value. From a repackaged PaaS layer to a revamped data stack, announcements at Google Cloud Next 2026 reinforce that this will be the company’s next chapter.
If Deloitte, EY, KPMG and PwC enable AI at scale within their own organizations, they should be able to successfully compete with firms like Grant Thornton, Protiviti and Kearney for consulting spend by companies in the $500 million to $5 billion range. In almost 19 years of watching the Big Four firms operate, I’ve seen countless small- and medium-sized enterprises’ initiatives launch, falter and fade. AI promises to upend that track record and, finally, make these firms players in the midmarket.
Except it won’t, at least not any time soon. For starters, anyone who has seen enterprisewide AI adoption at scale knows that success, when it comes, comes in small, incremental steps, not as massive, business-model-altering change. AI implementation is harder than it looks, and AI adoption at scale requires time, tech, leadership, experimentation and change management.
The Big Four firms provide exceptional advice on adopting AI at scale and have become adept at helping clients on their AI journeys (“tell me what to do” and “do it for me”), but they’re not immune to the challenges all large organizations face. In fact, given their consensus-dependent organizational model, these firms might face higher hurdles than the average top-down decision-making company.
Smaller engagements mean more clients, and although AI offers assistance, shouldn’t a midmarket client paying Big Four fees receive Big Four quality and service? Are the AI-enabled solutions equivalent to human expertise, and do they provide cost savings for everyone? TBR research says no to the latter, as digital full-time employees cost more than human workers, at least right now.
Until those midmarket companies aiming to become large global enterprises heed the advice of junior partners at Big Four firms, those partners are stuck cultivating, tending and harvesting more clients to close the revenue gap with senior partners. Do they get the chance to lead a major client like Citibank or be their firm’s global financial services lead by serving a local savings bank? No. Serious change management needs to happen within the Big Four, enabled by AI and specifically addressing the organizational, reputational, and compensational challenges of sustaining an investment in the midmarket. Do these firms have more on their plate right now? Yes. (Learn more in our Management Consulting Benchmark.)
The “tier two” firms listed in the previous section and their peers bring three strengths that help keep the Big Four at bay:
2026 Comcast Business Analyst Conference, Philadelphia, April 15-16, 2026 — A select group of industry analysts gathered at the Comcast Center in Philadelphia to hear from Comcast Business leaders about the progress and success of the unit’s sales and go-to-market strategies. The event continued to center on its theme introduced at last year’s conference, “Everything, Everywhere, All at Once,” reflecting the increasingly complex operating environment customers face and Comcast Business’ role in helping them navigate change through integrated solutions. Building on this theme, Comcast Business emphasized the accelerating pace of innovation over the past year, underscoring advancements in AI and network capabilities as it aims to deliver solutions that keep pace with the speed of business transformation. The event was hosted by NBC News Business and Data Correspondent Brian Cheung and included a State of the Business session with Comcast Business President Edward Zimmermann, a Strategy & Vision session with Comcast Business Chief Product Officer Bob Victor, and an update on Comcast’s network from Chief Network Officer Elad Nafshi. The agenda also featured panel discussions with senior leadership, speaker sessions with Comcast Business customers, and fireside chats with high-profile thought leaders on AI development and trends.
Since 2025, Comcast Business has accelerated its transition from a connectivity-led provider to a solutions- and platform-oriented partner for enterprise customers. The 2026 analyst conference highlighted the company’s focus on expanding share among global enterprises through continued investment in AI-enabled networking, cybersecurity and edge compute capabilities. This evolution reflects both opportunity and necessity. Enterprise growth is increasingly driving overall performance, while the SMB segment faces intensifying pricing pressure from fixed wireless access (FWA) and converged offerings.
At the same time, rapid advancements in AI are reshaping customer requirements, placing greater emphasis on low-latency connectivity, integrated security and real-time data processing. Comcast Business is positioning itself to capitalize on these trends by leveraging its network scale, partner ecosystem and managed services portfolio to deliver differentiated outcomes. However, success will depend on the company’s ability to execute, particularly whether it can monetize AI-driven capabilities and scale its global platform.
Comcast Business’ revenue performance remains relatively strong, generating over $10.2 billion in 2025, exceeding its long-term goal of reaching $10 billion in annual revenue. Growth is increasingly driven by the enterprise segment, which expanded 13.1% in 2025, supported by the integration of acquisitions, such as Masergy and Nitel. Additionally, the company now serves approximately 90% of Fortune 500 companies in some way. Comcast Business is also expanding its focus on multinational enterprises, leveraging partnerships with global operators across more than 130 countries.
Despite this momentum, the SMB segment — the company’s largest revenue contributor — is becoming increasingly challenging. Competition from FWA providers and converged offerings in the U.S. market is intensifying pricing pressure as small businesses gravitate toward lower-cost “good enough” connectivity solutions. These dynamics contributed to a net loss of 48,000 business customer relationships in 2025, compared to a net loss of 16,000 in 2024 and net additions of 17,000 in 2023. TBR believes the majority of these losses occurred within the SMB segment.
To offset customer losses, Comcast Business is increasing its focus on cross-selling value-added services to customers in areas such as mobility, SD-WAN, security and unified communications. For instance, Comcast Business reported that its enterprise customers are spending three times as much for value-added services as on core connectivity services compared to 2023. Comcast Business will also increase wireless revenue from larger businesses in 2026 through its new MVNO agreement with T-Mobile. The agreement covers up to 1,000 lines per account, which will enable Comcast to begin targeting the midmarket with wireless offerings, whereas its existing B2B MVNO agreement with Verizon is limited to 20 lines per account.
Comcast is expanding its use of AI from targeted, efficiency-driven applications to a more pervasive, embedded role across its network, solutions portfolio and customer engagement model. AI is now integrated across key areas, including network optimization, cybersecurity, sales enablement and customer experience, and is improving operational efficiency through internal use cases such as automated RFP development, deep research and meeting summarization. AI integration is enabling Comcast to automate over 99.7% of software changes across its network, supporting self-healing capabilities that can quickly resolve outages and, over time, help improve customer retention.
Comcast expects AI to not only enhance network and operational efficiencies but also create meaningful revenue-generation opportunities, though the company remains in the early stages of developing monetization strategies. For example, Comcast’s edge computing capabilities support ultra-low latency speeds of less than 1 millisecond for many customers, positioning the company to enable advanced AI-driven applications such as AR/VR, which are more dependent on low latency than text-based use cases. Comcast Business is also exploring customer-facing AI use cases, including small-business concierge agents designed to manage front-desk functions such as greeting customers, scheduling appointments and handling routine inquiries, highlighting the potential to extend AI-driven value beyond internal operations and into customer-facing revenue opportunities.
The company is advancing its enterprise strategy through the formal launch of Comcast Business Innovation Labs, an initiative designed to codevelop and rapidly scale first-to-market solutions for midmarket and enterprise customers. The lab brings together Comcast Business, customers and a broad ecosystem of technology partners to address specific business challenges, reflecting a more demand-driven approach to innovation. A key focus for Comcast Innovation Labs is supporting edge and AI-driven use cases by leveraging Comcast’s network capabilities and partner ecosystem.
Initial programs launched under the Comcast Business Innovation Lab include a partnership with Dell Technologies to deliver managed edge compute for AI and real-time applications and partnering with Digital Realty to enable seamless hybrid and multicloud connectivity through data center fabric services. Comcast Business is also collaborating with Expedient to support three core capabilities: AI operations at scale via Expedient’s Secure AI CTRL services, private cloud as a cost-efficient environment for workloads, and managed disaster recovery to support mission-critical applications.
TBR believes Comcast Innovation Labs strengthens the company’s ability to differentiate through ecosystem-driven innovation and faster solution development cycles, particularly as enterprise customers seek more tailored outcome-based offerings. However, the long-term impact of the initiative will depend on Comcast Business’ ability to scale these solutions beyond pilot environments and integrate them effectively across its broader portfolio and go-to-market strategy.
The 2026 Comcast Business Analyst Conference highlighted the company’s evolution from a connectivity-focused provider to a solutions-oriented partner for enterprise customers. Comcast Business’ ability to surpass $10 billion in annual revenue and sustain double-digit enterprise growth underscores the effectiveness of its upmarket strategy, supported by acquisitions, global partnerships and an expanding portfolio of value-added services.
However, SMB, which accounts for the majority of Comcast Business’ revenue, is becoming increasingly challenging as FWA competition and macroeconomic pressures drive greater pricing sensitivity. These headwinds will require Comcast Business to further strengthen its value proposition to retain and grow its SMB base and combat competitive pressures in the market.
NVIDIA’s increasing emphasis on physical AI signals that the company’s ambitions extend well beyond digital workloads. By linking its agent software stack with simulation, robotics and autonomous systems, NVIDIA is positioning itself as the foundational platform for both virtual and real-world AI applications. GTC 2025 established the importance of inference, and GTC 2026 clarified that the next phase of AI will be defined by agents, and NVIDIA is building the infrastructure to power them from end to end
This quarter, TBR FourCast looks at Accenture, Capgemini, HCLTech and IBM Consulting, comparing how their underlying data strategies, especially related to engineering and integration, prepares them for advanced AI adoption.
The U.S. federal government is the largest single buyer of IT services in the world, making it a critical customer target for leading providers in the space. For the current federal fiscal year (FFY), U.S. federal IT spending is estimated to approach $130 billion. Within that umbrella of spending, the Department of Defense (DOD) is not only the largest driver of spend but is also expected to see the most significant spending growth, at an estimated 5% year-to-year. Cloud-delivered options have been increasingly important to the DOD, most notably with the $9 billion Joint Warfighting Cloud Capability (JWCC) contract in 2022 and the newest iteration of the vehicle, dubbed JWCC Next.
The shift to cloud continues, but AI has become the clear priority for the DOD’s large and increasing IT investments over the past six months. As outlined in TBR’s 3Q25 Federal IT Services Benchmark: “TBR believes federal agencies increasingly view AI as an essential technology for enhancing mission workflows rather than as a niche, specialized tool or tool set. As such, we anticipate broadly accelerating implementation of comprehensive, agencywide AI platforms in FFY26 and FFY27. FSIs [federal systems integrators] will be tapped to not only integrate AI into IT infrastructures but also develop secure and ethically sound foundations for AI adoption.”
All of this is a long-winded way of setting up just how important the recent developments between Anthropic, OpenAI and Palantir are considering the implications for the largest agency (DOD), within the single largest buyer of IT in the world (U.S. Federal Government), relating to the single largest technology priority (AI).
Anthropic took a firm stance that the DOD could not use the company’s Claude technology for mass civilian surveillance or in fully autonomous weapons. This position caused Anthropic to lose the contract and receive a designation as a national security risk, threatening its partnerships with other providers. For Anthropic, the loss is not just the $200 million DOD agreement ceiling it won in July 2025.
On Feb. 12, Anthropic executives said the company’s run-rate revenue was $14 billion, and it raised $30 billion at a $380 billion valuation that same month. A few weeks later, Anthropic executives told a court the Pentagon blacklist could reduce 2026 revenue by multiple billions of dollars. Company leadership has also argued that the formal legal scope is narrower than the political rhetoric and that it should apply only to Claude’s use in direct DOD contract work, not all business with contractors. Reuters reported the Pentagon left room for exemptions in “rare and extraordinary circumstances.”
That means the real financial risk is probably not one canceled award but rather a pipeline contamination: Contractors derisking away from Claude, slower federal conversions, and reputational drag in defense-adjacent enterprise sales. Put differently, the $200 million ceiling is only about 1.4% of Anthropic’s disclosed $14 billion run rate, so the “multiple billions” warning has to be about second-order effects, not just the contract itself.
For OpenAI, the near-term revenue uplift is real but probably less dramatic than the strategic win. OpenAI’s federal posture was already building before Anthropic’s rupture. OpenAI launched and scaled usage of ChatGPT Gov, announced a $200 million-ceiling pilot with the DOD’s Chief Digital and Artificial Intelligence Office, struck a General Services Administration (GSA)-wide deal offering ChatGPT Enterprise to agencies for $1 per agency for a year, brought ChatGPT onto GenAI.mil for a platform used by 3 million civilian and military personnel, and most recently added an Amazon Web Services (AWS) route to sell models to U.S. defense and government agencies for classified and unclassified work.
Although the U.S. government activity is notable, it still represents a small portion of OpenAI’s overall revenue, which was rumored to have surpassed a $25 billion run rate as of early 2026. Put in this context, a $200 million government award represents only 0.8% of that run rate. However, the much bigger financial effect is strategic: Anthropic’s loss makes OpenAI the default frontier-model substitute for defense buyers, which should raise public-sector lifetime value, accelerate follow-on pipeline conversion, and strengthen valuation support.
The impact on Palantir of the change in DOD AI provider from Anthropic to OpenAI includes a very modest short-term financial upside and, more importantly, a reinforcement of Palantir’s position underpinning U.S. government AI technologies. Palantir clearly leans into DOD engagement and lacks any qualms about the use of its technology by the military and controversial domestic agencies like Immigration and Customs Enforcement. Palantir’s revenue is also much more highly dependent on the government sector; in 2025, $2.4 billion of Palantir’s total revenue, or roughly 53.7%, was generated by government contracts.
Palantir’s FedStart program is an on-ramp to absorb the federal compliance burden for other vendors on a usage basis, with Palantir handling ATO (Authority to Operate) conversations, compliance artifacts, continuous monitoring and control assessments. Anthropic joined FedStart in 2025, but Palantir integrates multimodal AI and its partners with Microsoft to operationalize Azure OpenAI in classified government environments.
With TBR Insight Center’s interactive data visualization tool, your team can quickly adapt the thousands of data points within the AI & GenAI Model Provider Market Landscape, Cloud Data & Analytics Market Landscape and U.S. Federal Cloud Ecosystem Report for tailored competitive analysis, go-to-market strategy and executive briefings. The tool enables you to curate relevant quantitative insights by company, business unit and/or market segment, creating a report specific to your needs and ensuring consistent frameworks across projects.
Click here to explore Insight Center’s data visualization tool, or start your free trial today to access this one-of-a-kind digital-first intelligence platform.
If you believe you have access to the full research via your employer’s enterprise license or would like to learn how to access the full research, click here.
Memory manufacturers continue to shift their capex investments toward expanding high-bandwidth memory (HBM) production capacity in support of rampant AI server demand. At the same time, demand for more commoditized dynamic random access memory (DRAM), such as DDR5 and LPDDR5X, is also growing but at a slower rate relative to HBM demand. The combination is driving a supply-and-demand imbalance in the DRAM market, leading to higher prices.
In the below TBR Insights Live session, Principal Analyst Angela Lambert and Senior Analyst Ben Carbonneau share insights into how rising memory prices and Windows PC ecosystem investments will impact PC refresh and the adoption of AI PCs in 2026 and beyond.
This TBR Insights Live session is available on demand on our YouTube channel. Visit this link to download the presentation’s slide deck.
If you’d like to further explore the data mentioned in this TBR Insights Live session, sign up for a free trial of TBR Insight Center™ today.
TBR Insights Live sessions are held typically on Thursdays at 1 p.m. ET and include a 15-minute Q&A session following the main presentation. Previous sessions can be viewed anytime on TBR’s Webinar Portal.