DOGE Federal IT Vendor Impact Series: IBM Federal

The Trump administration and its Department of Government Efficiency (DOGE) have generated massive upheaval across the board in federal operations, including in the federal IT segment. As of March 2025, thousands of contracts described by DOGE as “non-mission critical” have been canceled, including some across the federal IT and professional services landscape. TBR’s DOGE Federal IT Vendor Impact Series explores vendor-specific DOGE-related developments and impacts on earnings performance. Click here to receive upcoming series blogs in your inbox as soon as they’ve published.

 

DOGE’s aggressive cost-cutting activities impacted IBM-Fed* in 1Q25

IBM tendered its 1Q25 earnings on April 23, and while the company does not disclose fiscal data about the federal operations of IBM Consulting, IBM’s executives did provide useful color on IBM-Fed and the impact of DOGE. Not surprisingly, IBM-Fed’s contracts with the U.S. Agency for International Development (USAID), much maligned by the Trump administration, suffered cancellations and drawdowns.
 
According to the DOGE-Terminated Contracts Tracker on the GX2 website, which tracks developments in federal contracting, IBM-Fed has had a total of $40.1 million in contracts terminated by DOGE as of the publication of this blog. Cancellations included awards with the Department of the Treasury ($17.5 million in TCV), the Department of Health & Human Services ($3.4 million in TCV), the Commerce Department ($1.3 million in TCV), and the Department of Education ($18 million in TCV). Without disclosing specific revenue data for IBM-Fed, IBM noted that its federal business accounts for less than 5% of IBM’s total corporate revenue and less than 10% of IBM Consulting sales, or, according to TBR estimates, about $490 million in 1Q25, up 3% year-to-year.
 
We note that none of the USAID awards terminated or scaled back by DOGE were listed on the GX2 website. IBM CFO Jim Kavanaugh indicated during the 1Q25 earnings call that IBM-Fed had a “handful of contracts” canceled by DOGE, affecting about $100 million worth of contracts in IBM Consulting’s $30 billion (on an annualized basis) backlog.

The advisory business within IBM-Fed bore the brunt of DOGE-based pressures; the company’s core technology operations may have largely been spared

IBM indicated during the earnings call that 40% of IBM-Fed’s revenue stems from technology-focused work described as “high-value annuitized revenue under contract” and, by implication, is so far unscathed by DOGE. IBM-Fed blends its hybrid cloud, AI and security technologies to offer federal agencies a suite of transformative solutions that are very technology-centric and mission-enabling by nature. Conversely, 60% of IBM-Fed’s sales derive from advisory-based work, which company executives noted during the earnings call would be “more susceptible to discretionary efficiency-type programs.”
 
Based on data about IBM-Fed’s canceled contracts on the GX2 website, we believe the advisory work affected by DOGE included cloud transition and support services, data standards testing and implementation, data quality support services, the acquisition and implementation of integrated workplace management system licenses, and “data at rest” support services (i.e., data that is stored and not being actively used or transmitted). Other contracts were “terminated for convenience,” according to the GX2 website, which did not provide a specific description of the canceled services.
 
IBM-Fed, according to IBM CEO Arvind Krishna, processes claims for veterans, provides procurement services to the General Services Administration (GSA), and has implemented and is currently operating payroll systems for several federal agencies. Krishna acknowledged that “some areas around the edges” of this work “could be viewed as discretionary” by DOGE, but that the bulk of IBM-Fed’s services are mission critical and technology focused.

IBM-Fed will double down on its core cloud, security and AI capabilities to successfully traverse the DOGE-disrupted federal IT space in 2025

According to TBR’s 1Q25 IBM Consulting Earnings Response, “IBM Consulting could experience variability in revenue growth in 2025, and IBM is cautious about the revenue contribution from the business to total corporate revenue due to possible further tightening of discretionary spending driven by macroeconomic uncertainty and the U.S. Department of Government Efficiency’s (DOGE) activities.
 
However, IBM Consulting will continue to gain ground in areas such as generative AI (GenAI) due to IBM’s early advances in the segment, diversifying revenues through new areas of expansion.” To buffer its 2025 sales growth against DOGE’s cost-rationalization efforts and offset revenue losses from the cancellation of advisory work deemed discretionary (and thus expendable) by DOGE, IBM-Fed must play to its strengths in AI- and security-infused hybrid cloud solutions and emphasize how well its offerings align with DOGE’s efficiency agenda.
 
IBM-Fed won large-scale programs with civilian and defense agencies in 2024, thanks to the additional delivery and offerings scale in digital transformation it obtained by acquiring Octo Consulting in late 2022, another advantage and a key selling point for IBM-Fed when advising or coaching the DOGE advisory board. While Octo’s pure play advisory capabilities expose IBM-Fed to DOGE’s federal spending cuts in traditional consulting services, Octo’s oLabs center of excellence showcases IBM-Fed’s acquisition-enhanced cloud, security, data science and DevSecOps capabilities that sync well with the IT priorities of the Trump administration.

IBM-Fed must accelerate its expansion within the DOD and among national security agencies, particularly by emphasizing its strengths in cloud

Octo’s oLabs also serves national security and defense agencies. The Trump administration has indicated national security will be an overarching budget priority during its term and has hinted at a federal fiscal year 2026 (FFY26) defense budget surpassing $1 trillion for the first time, underscoring the urgency for IBM-Fed to accelerate its expansion with the Pentagon, where it has been gaining traction since acquiring Octo.
 
According to TBR’s 1H25 IBM Federal Vendor Profile, “Some federal IT industry observers believe the Trump administration’s DOGE will accelerate cloud investment as federal agencies may be forced to outsource more operations deemed outside ‘Inherently Governmental Functions (IGF).’ Cloud adoption in the Department of Defense (DOD) continues to far exceed civilian cloud investment, which the GSA’s Federal IT Dashboard (FITD) estimated to be $8.2 billion in FFY24, up from $5.5 billion in FFY23.”
 
IBM-Fed could leverage IBM’s 1Q25 $6.4 billion acquisition of HashiCorp to accelerate DOD-based expansion, as HashiCorp has helped the DOD migrate more than 3,000 applications to the cloud with its Terraform (Infrastructure as Code software) and Vault (identity-based security) tools designed to facilitate migrations to multicloud architectures. The DOD has clearly indicated it favors a multicloud approach for implementing cloud-based edge computing solutions.
 
*TBR refers to IBM Consulting’s federal IT operations as IBM-Fed. IBM-Fed is not an official business line title used by IBM or IBM Consulting. The business defined by TBR as IBM-Fed resides within IBM Consulting’s U.S. Public and Federal Market group.

 

TBR’s DOGE Federal IT Impact Series will include analysis of Accenture Federal Services, General Dynamics Technologies, CACI, IBM, CGI, Leidos, IFC International, Maximus, Booz Allen Hamilton and SAIC. Click here to download a preview of our federal IT research and receive upcoming series blogs in your inbox as soon as they’ve published.

 

DOGE Federal IT Vendor Impact Series: CACI

The Trump administration and its Department of Government Efficiency (DOGE) have generated massive upheaval across the board in federal operations, including in the federal IT segment. As of March 2025, thousands of contracts described by DOGE as “non-mission critical” have been canceled, including some across the federal IT and professional services landscape. TBR’s DOGE Federal IT Vendor Impact Series explores vendor-specific DOGE-related developments and impacts on earnings performance. Click here to receive upcoming series blogs in your inbox as soon as they’ve published.

 

CACI spared from major DOGE disruptions: Growth and profitability on track for FY25 goals

CACI tendered its 1Q25 earnings on April 23, and TBR did not discern any material impact from DOGE on the company’s business during the quarter, the third fiscal quarter of CACI’s FY25 (ending June 30). The company posted sales of $2.17 billion in 1Q25, up 11.8% year-to-year on a statutory basis and up 5.6% on an organic basis. CACI’s gross margin of 33.8% in 1Q25 was up sequentially from 33.2% in 4Q24, while its operating margin of 9.1% in 1Q25 was up 50 basis points sequentially from 8.6% in 4Q24. The company’s adjusted EBITDA margin was 11.7% in 1Q25, up from 11.1% in 4Q24.
 
CACI believes demand will remain strong through the remainder of its FY25 and into its FY26 for technologies and capabilities at the core of the company’s portfolio: AI-enhanced and commercially honed software-defined solutions delivered with Agile development methodologies; signals intelligence (SIGINT) and electronic warfare (EW) technologies for warfighters, defense vehicles and platforms, and IC applications; and AI-infused financial management offerings.
 
Uninterrupted sales growth and consistent margin performance indicate CACI’s offerings remain well aligned to the Trump administration’s IT investment priorities, particularly as the new administration prepares to expand investment in cybersecurity, national security and national defense, and advanced space-based communications systems for defense, intelligence and civil applications. CACI executives also noted that the federal budget environment is slowly becoming more constructive and more transparent, a positive harbinger for CACI and its fellow federal IT contractors.

CACI’s order book was essentially immune to DOGE-related turmoil in the federal IT market

TBR did not observe any impact from DOGE activities on CACI’s book of business. CACI’s backlog fell 1.3% sequentially, from $31.8 billion to $31.4 billion in 1Q25, but his kind of decline is typical in the company’s third fiscal quarter. CACI’s trailing 12-month (TTM) book-to-bill ratio was 1.5 in 1Q25, down from 1.7 in 4Q24. However, a sequential decline from the second to third fiscal quarter is not unusual for the company. In 1Q25, both the TTM book-to-bill ratio of 1.5 and the quarterly ratio of 1.2 were consistent with figures from the same period last year.
 
Furthermore, CACI’s bookings of $2.2 billion in 4Q23 and $3.5 billion in 1Q24 came during a period of exceptionally robust Department of Defense and Intelligence Community-related award activity. CACI’s bookings were $2.75 billion in 1Q25, up from $1.2 billion in 4Q24, consistent with the seasonal, historical pattern of sequential bookings expansion in the company’s third and second fiscal quarters. CACI noted in its 1Q25 earnings discussion that DOGE examined seven contracts in the company’s order book, including one that had already been completed. The aggregate revenue impact of these awards being eliminated by DOGE would only be $3 million in TCV, though DOGE has only notified CACI that $1 million worth of this ongoing work is likely to be canceled.
 
The company acknowledged that its business development teams have experienced some deceleration in certain aspects of the sales cycle, such as invoice and funding approvals. CACI CFO Jeffrey MacLauchlan said during the earnings call that “things that used to take two or three days are taking four or five days.” CACI’s leadership expects the disruption, which according to the company has been “very manageable” to date, to wane during the second half of federal fiscal year 2025 (FFY25). If sales motions are being impeded by DOGE, TBR would expect to see this reflected in lower-than-expected margin performance by CACI, but we did not observe any DOGE-related margin erosion in CACI’s P&L in 1Q25.

Undeterred by the DOGE-disrupted environment, CACI elevates several elements of its FY25 guidance

CACI raised the low end of its FY25 sales guidance range in 1Q25 and is now calling for top-line revenue of between $8.55 billion and $8.65 billion, implying a growth range of between 11.6% and 12.9% over FY24 revenue of $7.66 billion. In 4Q24 the company forecasted $8.45 billion in revenue at the low end of its projected FY25 sales range, implying growth of 10.3% at the bottom of the range.
 
CACI also raised the low end of its guidance for FY25 adjusted net income* in 1Q25 and now expects at least $543 million in FY25, up from $537 million forecasted in 4Q24.
 
CACI elevated its outlook for non-GAAP adjusted diluted earnings per share (ADEPS) in 1Q25, and as of 1Q25 is projecting a range of between $24.24 and $24.87 per share for FY25, up from a previous ADEPS range of between $23.24 and $24.13 per share. Free cash flow guidance was also elevated from $450 million tendered in 4Q24 to $465 million in 1Q25.
 
TBR notes that CACI has twice raised guidance for FY25 sales, adjusted net income, ADEPS and free cash flow since initially tendering its FY25 outlook in 2Q24. CACI is still guiding for a FY25 EBITDA margin in the low 11% range, implying a potential improvement of 100 basis points over FY24’s EBITDA margin of 10.4%, but also suggesting CACI does not expect any DOGE-related margin headwinds through the remainder of FY25.

CACI will remain vigilant and maintain a constant dialogue with customers

During CACI’s 1Q25 earnings call, CEO John Mengucci described DOGE’s objectives as “peace through strength, secure borders, increased efficiency and technology modernization.” Mengucci and his executive team remain confident that CACI’s strategy and portfolio are and will remain in sync with DOGE’s goals and with the IT strategy of the Trump administration, a contention supported by the company’s 1Q25 fiscal results and its more optimistic FY25 outlook.
 
Irrespective, CACI recognizes that federal executives are under pressure to accelerate IT modernization, quickly achieve IT-driven operational efficiencies and curb spending according to DOGE directives. Procurement teams at federal agencies are struggling to keep bid review processes and proposal adjudications on schedule as the Trump administration executes large-scale furloughs across the federal workforce. As such, CACI will keep its executives, business line leaders and business development teams as close as possible to IT decision makers and procurement counterparts in federal agencies for as long as DOGE’s efficiency agenda is in effect.
 
*Adjusted net income: GAAP-compliant net income excluding intangible amortization expense and the related tax impact

 

TBR’s DOGE Federal IT Impact Series will include analysis of Accenture Federal Services, General Dynamics Technologies, CACI, IBM, CGI, Leidos, IFC International, Maximus, Booz Allen Hamilton and SAIC. Click here to download a preview of our federal IT research and receive upcoming series blogs in your inbox as soon as they’ve published.

 

Fujitsu Expands AI Strategy in Europe, Emphasizing Collaboration, Compliance and Customization

‘AI can be a knowledge management accelerator, but only if well-fed by an enterprise’s own data’

In February TBR met with two AI leaders in Fujitsu’s European Platform Business to better understand the company’s approach to the AI market, its evolving AI capabilities and offerings, and what we can expect as 2025 unfolds. Maria Levina, AI Business analyst, and Karl Hausdorf, head of AI Business, gave a detailed presentation focused primarily on the European market. The following reflects both that briefing and TBR’s ongoing research and analysis of Fujitsu, the company’s partners and peers, and the overall AI landscape.
 
One highlight that illustrates many facets of Fujitsu’s approach to AI was Levina and Hausdorf’s description of Fujitsu’s customers’ choices between “bring your own data” and “bring your own AI.” The first allows more AI-mature customers to bring their data into a Fujitsu-provided on-premises solution with full support for scaling, maintaining hardware and software, and updating, as needed.
 
The second allows customers to run their own AI stack on a Fujitsu “validated and optimized” platform, developed and maintained by Fujitsu and select technology partners. Critical, in TBR’s view, is Fujitsu’s positioning of these options as responsive to clients’ needs, as determined by all AI stakeholders within an enterprise, including IT, AI and business leaders.
 
Levina and Hausdorf explained, “Together, with our ecosystem of partners, we’re committed to unlock the potential of generative AI for our clients” through “on-premises data sovereign, sustainable private GPT and AI solutions,” focused on “rapid ROI.” Fujitsu is not approaching clients with a technology solution, but rather with options on how to address and solve business problems. As the Fujitsu team said, “Understand the why, know the what, and co-create the how.” The Fujitsu team also noted that the company’s industry expertise resides in the processes and workflows unique and/or critical to an industry.
 

‘Maintaining control of data means owning your AI in the future’

Before diving into Fujitsu’s AI offerings, the Fujitsu team laid out their understanding of the European market, sharing data the company collected around AI adoption, use of AI platforms, and barriers to growth (in Fujitsu’s phrasing, “progress-limiting factors,” which is perhaps a more positive spin on the usual list of barriers and challenges). Fujitsu surveyed or spoke with 400 data and IT professionals across six European countries, and the results indicated that overcoming legacy mindsets continues to be a major impediment to adopting and harnessing the value of AI.
 
TBR’s November 2024 Voice of the Customer Research similarly noted challenges in Europe with “the lack of engagement from employees who are being asked to change.” The Fujitsu team noted that change management, therefore, had to involve all AI stakeholders, including “IT people, business people and AI people” within an enterprise.
 
In TBR’s experience, IT services companies and consultancies continue to find new constituents for change management at their clients as the promise — and disruption — of AI becomes more widespread, reinforcing Fujitsu’s strategy of bringing change management to all AI stakeholders. Lastly, the Fujitsu team noted that within European clients, expectations around AI have heightened, especially as AI initiatives have launched across multiple business units. Again, Fujitsu’s research and TBR’s Voice of the Customer Research align around ROI expectations as AI matures.
 
The Fujitsu team introduced their AI platform by delineating the key performance indicators they believe a successful platform must have: scaling, performance and speed, simplicity, energy efficiency, AI services in data centers, and GPUs.
 
Although TBR is not in a position to evaluate the technological strengths, completeness or complexity of Fujitsu’s platform, the expansive KPIs indicate Fujitsu has considered not only the IT needs behind an AI deployment but also the larger business factors, particularly the financial impacts. Levina and Hausdorf then dove into the details, including the two customer options described above (bring your own data and bring your own AI). They discussed how Fujitsu offers consulting around the technical and business implications of AI platforms and solutions, including an “AI Test Drive,” which allows clients to test AI solutions before investing in new technologies, large language models (LLMs) or other AI components.
 
Notably for TBR, Fujitsu’s presentation extensively highlighted the company’s AI alliance partners, including Intel, NVIDIA, AMD and NetApp, as well as a slew of LLM providers, demonstrating an appreciation for the collaborative and ecosystem-dependent nature of AI at the enterprise level. The Fujitsu team also stressed the European nature of its AI strategy and platform.
 
European clients, Fujitsu noted, had specific requirements related to the European Union’s (EU) General Data Protection Regulation and the EU AI Act, as well as a preference for on-premises solutions. Some of the use cases Levina and Hausdorf described included a law firm using Fujitsu-enabled AI solutions to analyze case data, contracts, corporate and public legal documents, and multiple deployments of Fujitsu-enabled private GPTs.

Additional observations

  • Fujitsu remains focused on targeting customers already aligned with the company around AI, a strategy that TBR believes speeds ROI and increases client retention.
  • In contrast to some peers in the IT services market, Fujitsu has capabilities across the entire AI technology stack — hardware, software and service — which Levina and Hausdorf called “highly appealing,” especially to European clients.
  • Levina and Hausdorf made two comments that, in TBR’s view, neatly sum up AI at present: “AI can be a knowledge management accelerator, but only if well-fed by an enterprise’s own data” and “maintaining control of data means owning your AI in the future.”

Fujitsu’s AI prowess makes it an invaluable partner

TBR has reported extensively on Fujitsu’s evolving AI capabilities and offerings, noting in a special report in May 2024: “TBR appreciates that Fujitsu’s combination of compute power and proven AI expertise makes the company a significant competitor and/or alliance partner for nearly every player fighting to turn GenAI [generative AI] hype into revenue.
 
“Second, Fujitsu’s vision of ‘converging technologies’ aligns exceptionally well with the more tectonic trends TBR has been observing in the technology space, indicating that Fujitsu’s market positioning is more strategic than transactional or opportunistic.” Add in Fujitsu’s deepening experience in delivering AI solutions to AI clients, and TBR continues to see tremendous near-term opportunity and growth for Fujitsu and its ecosystem partners.

Inside the AI Hardware Shift: Market Trends Every IT Decision Maker Should Watch in 2025

Register for Inside the AI Hardware Shift: Market Trends Every IT Decision Maker Should Watch in 2025

 

Silicon vendors and OEMs working together to support AI adoption

While OEMs are responsible for developing and delivering AI-driven and AI-enabling hardware offerings to market, silicon vendors’ innovations are at the heart of the AI hardware revolution.
 
The first wave of AI hardware demand has centered on high-performance AI infrastructure purpose-built to support large-scale AI model training workloads. But the rise of AI inferencing is giving way to a second wave of AI hardware demand as clients increasingly transition from the prototyping phase to the deployment phase with custom AI solutions. On the infrastructure side of the AI hardware market, OEMs such as Dell Technologies, Hewlett Packard Enterprise and Supermicro are integrating accelerated computing platforms from companies like NVIDIA. On the client devices side of the market, OEMs such as HP Inc. and Lenovo are developing new AI PC offerings based on system on a chip (SoC) platforms developed by AMD, Intel, Qualcomm and the like.
 
Join Senior Analyst Ben Carbonneau and Principal Analyst Angela Lambert Thursday, May 22, 2025, for an update on developments within the rapidly expanding AI PC and AI server markets as well as key findings from TBR’s AI PC and AI Server Market Landscape. This new research explores the nuances and interconnectedness of the semiconductor and OEM hardware industries, comparing market shares across various industry views and highlighting competitive analysis and forward-looking insights.

In this free session on AI hardware market trends you’ll learn:

  • TBR’s forecast for the AI PC and AI PC SoC markets
  • Our performance outlook for the AI server and AI server GPGPU (general-purpose computing on GPUs) markets
  • The latest industry trends and ecosystem partnerships
  • Key market dynamics contributing to and inhibiting growth

Register Now

 
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.
 

Google Cloud Cements Values of Enterprise Readiness, Full-stack AI and Hybrid Cloud at Next 2025

In April Google Cloud hosted its annual Next event to showcase new innovations in AI. Staying true to the theme of “A New Way to Cloud,” Google focused on AI, including how AI can integrate with enterprises’ existing tech landscape, with partners playing the role of orchestrator. After Google CEO Sundar Pichai spoke about the company’s achievements around Gemini, which is integral to Google Cloud’s strategy, Google Cloud CEO Thomas Kurian highlighted the business’s three key attributes: optimized for AI; open and multicloud; and enterprise-ready. Additionally, Google Cloud announced a series of new innovations that highlight how the company is trying to execute on these three areas to be the leader in modern AI development.

Google takes an end-to-end approach to AI

When discussing Google Cloud’s three key attributes, Kurian first highlighted how Google Cloud Platform (GCP) is optimized for AI. Based on our own conversations with IT decision makers, this claim is valid: many customers enlist GCP services purely for functional purposes, as they believe they cannot obtain the same performance with another vendor. This is particularly true of BigQuery, for large-scale data processing and analytics, and increasingly Vertex AI, which now supports over 200 curated foundation models for developers.
 
Within this set of models is, of course, Gemini, Google’s own suite of models, including the new Gemini 2.5 Pro, which has a context window of 1 million tokens and is reportedly now capable of handling advanced reasoning. To be fair, Google still faces stiff competition from other frontier model providers, but Google’s years of AI research through DeepMind and its ability to have models grounded in popular apps like Google Maps, not to mention Google Search, will remain among its key differentiators.
 
With that said, the AI software stack is only as effective as the hardware it runs on. That is why Google has been making some advances in its own custom AI accelerators, and at the event, Google reaffirmed its plans to invest $75 billion in total capex for 2025, despite the current macroeconomic challenges. A large piece of this investment will likely focus on paying for the ramp-up of Google’s sixth-generation TPU (Tensor Processing Unit) — Trillium — which became generally available to Google Cloud customers in December. Additionally, Google is making some big bets on the next wave of AI usage: inference.
 
At the event, Google introduced its seventh-generation TPU, dubbed Ironwood, which reportedly scales up to 9,216 liquid cooling chips linked through a high-powered networking layer, to support the compute-intensive requirements of inference workloads, including proactive AI agents. In 2024 there was a 3x increase in the number of collective TPU and GPU hours consumed by GCP customers, and while this was likely off a small number of hours to begin with, it is clear that customers’ needs and expectations around AI are increasing. These investments in AI hardware help round out key areas of Google’s AI portfolio ― beyond just the developer tools and proprietary Gemini models ― as part of a cohesive, end-to-end approach.
 

Watch now: Cloud market growth will slow in 2025, but will activity follow? Deep dive into generative AI’s impact on the cloud market in 2025 in the below TBR Insights Live session

 

Recognizing the rise of AI inference, Google Cloud reinforces longtime company values of openness and hybrid cloud

With its ties to Kubernetes and multicloud editions of key services like BigQuery and AlloyDB, Google Cloud has long positioned itself as a more open cloud compared to its competitors. However, in recent quarters, the company has seemed to hone this focus more closely, particularly with GDC (Google Distributed Cloud), which is essentially a manifestation of Anthos, Google’s Kubernetes-based control plane that can run in any environment, including at the edge. GDC has been the source of some big wins recently for Google Cloud, including with McDonald’s, which is deploying GDC to thousands of restaurant locations, as well as several international governments running GDC as air-gapped deployments.
 
At Next 2025, Google announced it is making Gemini available on GDC as part of a vision to bring AI to environments outside the central cloud. In our view, this announcement is extremely telling of Google Cloud’s plans to capture the inference opportunity. Per our best estimate, roughly 85% of AI’s usage right now is focused on training, with just 15% in inference, but the inverse could be true in the not-too-distant future. Not only that, but inference will also likely happen in distributed locations for purposes of latency and scale. Letting customers take advantage of Gemini to build applications on GDC — powered by NVIDIA Blackwell GPUs — on premises or at the edge certainly aligns with market trends and will help Google Cloud ensure its services play a role in customers’ AI inference workloads regardless of where they are run.

Boosting enterprise mindshare with security, interoperability and Google-quality search

Kurian mentioned that customers leverage Google Cloud because it is enterprise-ready. In our research, we have found that while Google Cloud is highly compelling for AI and analytics workloads, customers believe the company lacks enterprise-grade capabilities, particularly when compared to Microsoft and Amazon Web Services (AWS). But we believe this perception is changing, and Google Cloud is recognizing that to gain mindshare in the enterprise space, it needs to lead with assets that will work well with customers’ existing IT estates and do so in a secure way. This is why the pending acquisition of Wiz is so important. As highlighted in a recent TBR special report, core Wiz attributes include not only being born in the cloud and able to handle security in a modern way but also connecting to all the leading hyperscalers, as well as legacy infrastructure, such as VMware.
 
Google Cloud has been very clear that it will not disrupt the company’s multihybrid capability. In fact, Google Cloud wants to integrate this value proposition, which suggests Google recognizes its place in the cloud market and the fragmented reality of large enterprises’ IT estates. Onboarding Wiz, which is used by roughly half of the Fortune 500, as a hybrid-multicloud solution could play a sizable role in helping Google Cloud assert itself in more enterprise scenarios. In the meantime, Google Cloud is taking steps to unify disparate assets in the security portfolio.
 
At Next 2025, Google Cloud launched Google Unified Security, which effectively brings Google Threat Intelligence, Security Operations, Security Command Center, Chrome Enterprise and Mandiant into a single platform. By delivering more integrated product experiences, Google helps address clients’ growing preference for “one hand to shake” when it comes to security and lays a more robust foundation for security agents powered by Gemini, such as the alert triage agent within Google Security Operations and the malware analysis agent in Google Threat Intelligence to help determine if code is safe or harmful.
 
One of the other compelling aspects of Google’s enterprise strategy is Agentspace. Launched last year, Agentspace acts as a hub for AI agents that uses Gemini’s multimodal search capabilities to pull information from different storage applications (e.g., Google Drive, Box, SharePoint) and automate common productivity tasks like crafting emails and scheduling meetings. At the event, Google announced that Agentspace is integrated with Chrome, allowing Agentspace users to ask questions about their existing data directly through a search in Chrome. This is another clear example of where Google’s search capabilities come into play and is telling of how Google plans to use Agentspace to democratize agentic AI within the enterprise.

Training and more sales alignment are at the forefront of Google Cloud’s partner priorities

Google Cloud has long maintained a partner-first approach. Attaching partner services on virtually all deals; taking an industry-first approach to AI, particularly in retail and healthcare; and driving more ISV coselling via the Google Cloud Marketplace are a few examples. At Next 2025, Google continued to reaffirm its commitment to partners, implying there will be more alignment between field sales and partners, to ensure customers are matched with the right ISV or global systems integrator (GSI), a strategy many other cloud providers have tried to employ.
 
When it comes to the crucial aspect of training, partners clearly see the role Google Cloud plays in AI, and some of the company’s largest services partners, including Accenture, Cognizant, Capgemini, PwC, Deloitte, KPMG, McKinsey & Co., Kyndryl and HCLTech, have collectively committed to training 200,000 individuals on Google Cloud’s AI technology. Google has invested $100 million in partner training over the past four years, and as highlighted in TBR’s Voice of the Partner research, one of the leading criteria services vendors look for in a cloud partner is the willingness to invest in training and developing certified resources.

Google Cloud wants partners to be the AI agent orchestrators

As previously mentioned, Vertex AI is a key component of Google Cloud’s AI software stack. At Next 2025, Google Cloud introduced a new feature in Vertex called the Agent Development Kit, which is an open-source framework for building multistep agents. Google Cloud is taking steps to ensure these agents can be seamlessly connected regardless of the underlying framework, such as launching Agent2Agent (A2A), which is an open protocol, similar to protocols introduced by model providers like Anthropic.
 
Nearly all of the previously mentioned GSIs, in addition to Boston Consulting Group (BCG), Tata Consultancy Services (TCS) and Wipro, have contributed to the protocol and will be supporting implementations. This broad participation underscores the recognition that AI agents will have a substantial impact on the ecosystem.
 
New use cases will continue to emerge where agents are interacting with one another, not only internally but also across third-party systems and vendors. With the launch of the Agent Development Kit and the related protocol, Google Cloud seems to recognize where agentic AI is headed, and for Google Cloud’s alliance partners, this is an opportune time to ensure they have a solid understanding of multiparty alliance structures and are positioned to scale beyond one-to-one partnerships.

Final thoughts

At Next 2025, Google reportedly announced over 200 new innovations and features, but developments in high-powered compute, hybrid cloud and security, in addition to ongoing support for partners, are particularly telling of the company’s plans to capture more AI workloads within the large enterprise. Taking an end-to-end approach to AI, from custom accelerators to a diverse developer stack that will let customers build their own AI agents for autonomous work, is how Google Cloud aims to protect its already strong position in the market and help lead the shift toward AI inferencing.
 
At the same time, Google Cloud appears to recognize its No. 3 position in the cloud market, significantly lagging behind AWS and Microsoft, which are getting closer to each other in IaaS & PaaS revenue. As such, taking a more active stance on interoperability to ensure AI can work within a customer’s existing IT estate, and guaranteeing partners that have the enterprise relationships are the ones to orchestrate that AI, will help Google Cloud chart its path forward.

Trade Wars and the Professional Services Fallout: Talent, Growth and Operational Models in Flux

Significant market disruption likely in near and long term

Trade wars and tariff uncertainties conjure up visions of cargo ships, ports, factories and stacks of goods stranded by economic chaos, not consultants and IT services professionals. Fear, uncertainty and doubt are usually good for the consulting business, while the higher costs of running a business fuel demand for more outsourcing. This time, things might be different. This trade war, even if partially suspended for now, may significantly disrupt professional services, especially if tariffs continue creeping into new areas and the trust deficit continues to grow. Steel now, services later.
 
TBR believes three areas will likely experience added near-term stress if the trade war continues: acquisitions, sales cycles and staffing. Longer-term, more seismic changes may come to the H-1B visa program, regionalization efforts among the Big Four firms, and onshore/offshore talent models. Looming over all of these disruptions, at least at the moment, is the potential for a grand decoupling of the U.S. and China economies, with incomprehensible knock-on effects. Those near-term disruptions share a common denominator: macroeconomic uncertainty.
 
Making the business case for a significant acquisition becomes harder in a recession-fearing market. When clients extend sales cycles because they’re afraid to commit suddenly more precious resources to upgrades, modernizations or transformations, growth slows for consultancies and IT services companies. And when growth slows, so does hiring.
 
At its core, professional services is all about people. And when recruiting, rewarding and retaining people are pressured, everything is pressured. To understand how tariffs and trade wars could hurt consultancies and IT services companies, even in the short run, it is critical to step back and realize these professional services providers serve every industry. They may be in and of one industry themselves — professional services — but their clients span every industry that exists. When the steel, computer chip, automobile, bourbon and lumber industries get upended by tariffs, so do the consultancies and IT services companies serving them.
 

In 2025 IT services companies and consultancies will refine their alliances, articulate a clear joint value proposition, and align at both the leadership and salesforce levels. The most successful IT services companies and consultancies will be the ones that partner best. Learn more in TBR’s 2025 Ecosystems & Alliances Predictions special report.


 

Local and regional talent may be key to revenue growth

Powering through the near-term challenges, IT services companies and consultancies may then face structural changes to their operating environment, many centered on talent, starting with a reevaluation of the onshore/offshore mix. India-centric companies, which have historically relied on H-1B visas (at least to some degree; TBR appreciates that their reliance has varied widely), may find a less accommodating atmosphere in the U.S. and possibly even an unwillingness by potential candidates to relocate to the U.S.
 
At the same time, the Big Four firms may slow down their regionalization efforts, as having highly country-specific capabilities and dedicated staff may become a greater asset than more explicitly globalized organizations. TBR believes the more extreme outcomes around H-1B visas remain unlikely, while staying cognizant that the current trade war and tariff uncertainty also seemed unlikely a year ago. TBR does believe one highly likely outcome of the current trade crisis is a reassessment — by all IT services companies and global consultancies — of the overall onshore/offshore model. The recent uptick in global captive centers in India may be indicative of an enterprise trend toward more tightly owned and controlled offshore resources, but that was already the norm among IT services companies and consultancies prior to the trade war threat.
 
If trade wars persist, local and regional talent may become the key to sustained revenue growth, tied to local and regional economic growth overall. In other words, whichever company has the most and the best people on the ground in the fastest-growing places will continue to grow the most rapidly. It seems like a good time for the Big Four to have every country member firm run its own show as the on-the-ground market conditions start becoming even more disparate.
 

Watch now: TBR Principal Analyst & Practice Manager Patrick M. Heffernan discusses trend expectations for GenAI in the Professional Services market in 2025

Tariffs on services could further complicate market landscape

Returning to the starting image, trade wars evoke cargo ships, not consultants, and so far the Trump administration has not included services on the various tariff schedules. The U.S. currently runs a services trade surplus, and tariffs on services (as well as software) for various countries would be insanely difficult to assess. Artificial intelligence and the application of generative AI (GenAI) to procurement could make tariffs on services more manageable, but any efficiencies gained through those efforts would potentially erode the low-cost arbitrage advantage enjoyed by IT services companies and technology providers, damaging the overall U.S. trade balance.
 
Further complicating this picture, advances in AI and automation could mean any manufacturing jobs created in the U.S. as a direct result of tariffs would be digital FTEs, benefiting technology companies but undermining the Trump administration’s stated goals. In all, a mess, even if services remain off the tariff schedule.

Companies pursue multiple strategies around U.S.-China decoupling

Another potential scenario: Some economic and consulting leaders have been advocating for a U.S.-China decoupling for a few years, a possibility that is more likely now as every day brings another parry in the U.S.-China trade war. Some global consultancies have been kicked out of China. Others have downgraded their offices or quietly left on their own. And some are maintaining an arm’s-length relationship, and some are doing business as usual. Fools would predict which strategies will win out. TBR simply notes that companies may pursue multiple strategies.
 
For example, in August 2024 IBM closed its China Development Lab and China Systems Lab, laying off more than 1,000 employees across Beijing, Shanghai and Dalian. The closure was part of IBM’s initiative to relocate R&D functions to India and other countries due to competition and geopolitical tensions. However, IBM remains committed to working with clients in the Greater China region. In March IBM launched an initiative to expand in enterprise AI, hybrid cloud and industry-specific consulting services to drive digital transformation and implement AI and cloud solutions in China. As part of this initiative, IBM is working with China-based Great Wall Motor Co. Ltd. on digital transformation and global expansion. A complete decoupling may be unlikely, but consultancies and IT services companies that have financial flexibility and leaders who are prepared to take risks and withstand uncertainty will likely continue to thrive.

Here Comes KPMG: Client Trust, Alliance Focus and Tech-enabled Strategy Emphasized at 2025 Global Analyst Summit

Executing on its Collective Strategy through integrated scale and backed by robust strategic partnerships and platform-enabled services positions KPMG to remain a formidable competitor in the transforming professional services market

KPMG Global Chairman and CEO Bill Thomas kicked off the firm’s 2025 Global Analyst Summit by reinforcing the firm’s mission to be “the most trusted and trustworthy professional services firm.” As we have discussed at length across TBR’s professional and IT services research, firms like KPMG trade on trust with clients, alliance partners and employees. Putting a stake in the ground from the get-go provided Thomas and KPMG’s executives a strong foundation to rely on during the next two days as trust — at the human and technology level — was an underlying theme during presentations and demos.
As a member of the Big Four, KPMG has brand permission and a breadth of services that are relevant to nearly every role in any enterprise. As the firm executes on its Collective Strategy, TBR believes KPMG will accelerate the scale and completeness of its offerings, building on a solid foundation and expanding the gaps between KPMG and other consulting-led, technology-enabled professional services providers. ​
 
KPMG’s global solutions — Connected, Powered, Trusted and Elevate — which resonate with clients and technology partners, have now been brought together into one transformation framework under KPMG Velocity, providing KPMG’s professionals with clear insight into the firm’s strengths and strategy, and underpinning, in the near future, all KPMG’s transformation engagements. KPMG Velocity’s evolving strategy will challenge KPMG’s leaders to execute on the promise of that transformation during the next wave of macroeconomic pressures, talent management battles and technology revolutions. At the same time, KPMG’s leaders recognize that their priorities are transforming the firm’s go-to-market approach, unlocking the power of the firm’s people, reimagining ways of working, and innovating capabilities and service enhancements. ​
 
Success in executing these priorities, in TBR’s view, will come as KPMG shifts from building a foundation to scaling alongside the growing needs of its clients and as the era of GenAI presents yet another opportunity and challenge. Striking the right balance between elevating the potential of GenAI as a value creator and accounting for commercial and pricing model implications will test the durability of KPMG’s engagement and delivery frameworks. ​
 
Although the firm has placed in motion many of the aforementioned investments over the past 12 to 18 months, the one opportunity that is changing relates to speed. As one enterprise buyer recently explained to TBR: “GenAI will force all services vendors to change. The [ones] who [will] be [the] most successful will be [those] who do it fast.” With speed comes risk — which KPMG fully acknowledges and is why KPMG Velocity’s offering is a differentiator for the firm in the market. With KPMG Velocity, all of KPMG’s multidisciplinary and heritage risk and regulatory considerations have been embedded across each transformation journey to ensure clients can remain compliant and avoid the pitfalls that can often arise during transformation. ​
Continuing the firm’s presentation, Thomas outlined KPMG’s evolving Collective Strategy, noting that the firm is 18 months into its latest iteration focused on “accelerating trust and growth.” Among the key enablers of achieving this goal is KPMG’s collapsing of its organizational structure from 150 country-specific member firms to a cluster of 30 to 40 regionally organized “economic units.” TBR views this pivot as the most natural evolution of KPMG’s operating model. For the Big Four, the biggest challenge is how to demonstrate value through integrated scale. Once completed, the reorganization will allow KPMG to minimize such disruption and better compete for globally sourced opportunities from what the firm calls “transactions to transformation” and for large, multi-year, geographically dispersed enterprise, function and foundational transformations.
 
Following Thomas’ presentation, Carl Carande, KPMG U.S. & global lead, Advisory, and Regina Mayor, global head of Clients & Markets, amplified KPMG’s strategy, reinforcing the importance of the firm’s people, technology partners and technology — with AI the catalyst and change agent of success. For example, Carande recognized the technology relationships are changing in two ways. Relationships are becoming more exclusive, and the multipartner alliance framework offers a multiplier power — themes TBR has discussed at length throughout our Ecosystem Intelligence research stream.
 
Although KPMG continues to manage a robust network of alliance partners, highlighting its seven strategic partners — Google Cloud, Microsoft, Oracle, Salesforce, SAP, ServiceNow and Workday — solidifies its recognition of these vendors’ position throughout the ecosystem. Mayor expanded on Carande’s discussion around alliances through an industry lens, describing “alliance partners leaning in with KPMG” as they realize efforts to only sell the product will be insufficient. Meanwhile, on the KPMG side, alliance sales partners help figure out how to penetrate sector-specific alliance relationships.
 
Taking such a systematic approach across KPMG’s 7 sectors (with the desire to expand these to 14) will allow the firm to demonstrate value and support its evolving Collective Strategy to act as a globally integrated firm. Additionally, new offerings like KPMG Velocity (discussed in depth on Slide 6) will arm KPMG’s consultants with the necessary collective knowledge management to serve global clients locally, further supporting the firm’s strategy.
 
One could argue that many of KPMG’s steps, including launching partner-enabled industry IP, reinforcing trust, developing regionally organized operations, outlining a select few strategic partners, and investing in platform-enabled service delivery capabilities, resonate with the moves taken by many of its Big Four and large IT services peers. We see two differences.
 
First, KPMG is laser-focused on exactly which of the strategies above to amplify, rather than taking a trial-and-error approach. Second, it is about timing. Some of KPMG’s peers have tried these strategies for some time, with limited success because of poor execution or timing. We believe that as the professional services market goes through its once-in-a-century transformation, KPMG has an opportunity to ride the wave, provided it maintains internal consensus and executes on its operational and commercial model evolution with minimal disruption.
 

 

KPMG’s evolution will largely stem from orchestrating alliances with seven strategic technology partners

At the event, KPMG asserted the role of tech alliance partners in building the “firm of the future.” Although the firm works with a range of ISVs, a targeted focus on the firm’s seven strategic technology partners has become key to the company’s growth profile — with 50% of its consulting business alliance-enabled in the U.S. — and, as the case of previous audit client SAP shows, KPMG has been able to overcome barriers to ultimately help clients get the most out of technology. The firm’s approach of leading with client outcomes first and technology second is unchanged, but prioritizing a tech-enabled go-to-market approach will support KPMG’s position in the market behind two major trends.
 
The first trend is the overall maturation in partner alliance structures we see from the cloud vendors. Changes in programmatic structure, including bringing sales and partner delivery closer together, and an all-around shift in how partners are viewed among historically siloed vendors, could act as enablers for KPMG’s newer capabilities, including Velocity. Second, there is a big paradigm shift underway on where the value of tech exists. Increasingly, we see the firm moving down the stack, a trend enabled by agentic AI and customers’ need to harness their own data and build new applications. Across the Big Seven, there is no shortage of innovation. As the value of AI shifts down the technology stack, KPMG can leverage the technology to deliver business outcomes to clients.
 
To fully describe KPMG’s evolving technology alliance strategy and the firm’s growing capabilities, KPMG leaders hosted a panel discussion that included leaders from Microsoft, SAP, Salesforce and KPMG clients. Todd Lohr, KPMG’s head of Ecosystems for Advisory, set the stage by saying the firm views ecosystems as more than simply a collection of one-to-one alliances, but ecosystems are, instead, many-to-many relationships, an idea TBR has increasingly heard expressed by consultancies, IT services companies, hyperscalers and software vendors.
 
Having leaders from technology partners on stage to display a very common example of a tech stack — with SAP as the system of record (SOR), Salesforce in the front office, and Microsoft as the platform with Copilot — was a strong way to depict the “many-to-many relationships” structure and KPMG’s role in orchestrating the ecosystem, especially in scenarios where some of these ISVs may not have a native integration and/or formal collaboration with one another. Lohr noted that KPMG “needs to show up understanding how complicated multiparty relationships work before showing up and working them out ad hoc at the client.” That direct acknowledgment of the challenges inherent in multiparty alliances is decidedly not something TBR consistently hears from KPMG’s peers and partners.

KPMG moves away from vendor agnosticism

One of the most important takeaways for TBR from the summit was KPMG’s willingness, in the right circumstances, to aggressively abandon the typical agnostic approach to recommending technologies and instead make a specific technology recommendation where there is a deep understanding of the client needs. One client example highlighted this new(ish) approach. When the client reached out for advice on a sales-enablement platform, KPMG did not take an agnostic approach and, instead, told the client Salesforce was the only choice, based on KPMG’s evaluation.
 
Part of KPMG’s proposal rested on reworking the client’s processes so Salesforce could work as much out of the box as possible, limiting costs and customizations. As KPMG leaders described it, this reflected the opposite of most consultancies’ (and enterprises’) usual approach of forcing the business processes to work with a new technology. In a competitive bidding process, the lead KPMG partner, according to the client, answered questions on the Salesforce software and implementation issues without turning to others on the KPMG team, demonstrating mastery of Salesforce and the client’s IT environment that reassured the client about KPMG’s recommendations. Further, the client expressly did not want customization layers on top of Salesforce, knowing that would be more expensive over time.
 
Notably, the “fairly comprehensive implementation,” according to the client, took less than a year, including what the client said was “a lot of investment with KPMG in change management.” Recalling best practices TBR has heard in other engagements, the client team and KPMG called the Salesforce implementation Project Leap Frogs to avoid the word “transformation,” enabled champions across the enterprise, and held firm to the approach of making minimal customizations. In discussions with TBR, KPMG leaders confirmed that not being technology agnostic was contrary to the firm’s usual practice but was becoming more common.
 
Reinforcing that notion, a KPMG leader told TBR that the firm had lost a deal after it recommended Oracle and said SAP was not the right fit. The client selected SAP (for nontechnical reasons) but later awarded, without a competitive bidding process, Oracle-specific work to KPMG after noting respect for the firm’s honesty and integrity.

KPMG showcases client-centric innovation in action

ServiceNow implementation

A client story featuring a ServiceNow implementation that brought cost savings and efficiencies to the client notably emphasized change management, a core KPMG consulting capability that is sometimes overshadowed by technologies. The client described the “really good change management program that KPMG brought” as well as the emphasis on a clear data and technology core, out-of-the-box ServiceNow implementation, and limited customizations. In TBR’s view, KPMG’s approach with this engagement likely benefited considerably from the firm’s decades-long relationship with the client, playing to one of KPMG’s strengths, which the firm’s leaders returned to repeatedly in discussions with TBR: Trusted partnerships with clients create long-standing relationships and client loyalty.

Reimagining leaders

One client story centered on a five-day “reimagining leaders” engagement at the Lakehouse facility, conducted by the KPMG Ignition team. Surprisingly, KPMG included an immersive session with an unrelated KPMG team working on an unrelated client’s project that had little overlap with the business or technology needs of the leadership engagement client.
 
According to the KPMG Ignition team, the firm showcased how KPMG works, how innovation occurs at the working level, and how KPMG creates with clients, giving them confidence in KPMG’s breadth and depth of capabilities. Echoing sentiments TBR has heard during more than a decade of visiting transformation and innovation centers, KPMG Ignition leaders said that being enclosed on the Lakehouse campus made it easier for clients to be fully present throughout the engagement and removed from the distractions of day-to-day work.
 
KPMG kept the client in the dark about what to expect from the engagement, which prevented any biased expectations from creeping in before the engagement had even started. KPMG Ignition leaders shared additional insights, noting that it was a pilot program for rising leaders at the client, providing an immersive experience that showcased the power of the KPMG partnership.
 
Throughout the five-day immersion at KPMG Lakehouse, participants learned how to apply the methodologies that fuel innovation at KPMG while staying focused on one theme: reimagining leadership of the overall company and of the participants as next-generation leaders, as well as reimagining leadership capabilities at every level of the organization.
 
KPMG equipped the client’s leaders with methodologies emphasizing storytelling, design thinking and strategic insights, and strengthened the client’s culture by fostering high-performing, collaborative teams.
 
One final comment from the Ignition Center leaders: This pilot program “highlighted the fact that AI can be viewed as a wellness play across the agency if you free up capacity and understand what can be achieved.” Based on the use case and sidebar discussions TBR had with KPMG Ignition leaders, we believe Ignition Centers continue to evolve, although the basics remain the same: Get clients into a dedicated space outside their own office, use design thinking, and focus on business and innovation and leadership and change, not on technology.

The art of the possible

A final client story, presented on the main stage, wove together the themes of AI, transformation and trust. The client, a chemicals manufacturer and retailer, said KPMG consistently shared “what’s possible,” essentially making innovation an ongoing effort, not a one-off aspect of the relationship.
 
The client added that his company and KPMG had “shared values … and we understand each others’ cultures,” in part reinforced by KPMG dedicating the same team to the client during a multiyear engagement.
 
In TBR’s view, KPMG’s decision to highlight this client reinforced everything KPMG leaders had been saying during the summit: Relationships, built on consistent delivery and continually coming to the client with ideas and innovations, plus a commitment to the teaming aspect of the engagement, are KPMG’s superpower. Notably, this client was not a flashy tech company, a massive financial institution or a well-known government agency, and the work KPMG did was not cutting-edge or highly specialized but rather core KPMG capabilities — in short, what KPMG does well.

Velocity and GenAI: KPMG’s client-first approach to AI adoption and transformation

KPMG dedicated the second day of the analyst summit to AI, a decision that reflected the firm’s overall approach: Business decisions come first, enabled by technology. Supporting the firm’s AI strategy, KPMG has developed Velocity, a knowledge platform, AI recommendation and support engine underpinned by one universal method that pulls together every capability, offering and resource across the firm for the KPMG workforce. According to KPMG leaders, Velocity reinforces the firm’s multidisciplinary model and will become the primary way KPMG brings itself to clients.

In addition to sharing knowledge across the global firm, Velocity will help KPMG’s clients find the right AI journey that matches their ambitions — whether it be Enterprise, Function or Foundation — by allowing them to select a strategic objective they are trying to achieve, which function(s) they want to transform, and which technology platforms they want to transform on. The platform also reaffirms the firm’s acknowledgment of data’s role in AI. In fact, part of the rationale for Velocity was bringing the data modernization and AI business together while maintaining a focus on a sole client outcome. This means KPMG does not care whether customers build their data foundation with a hyperscaler or internally; as one leader in the AI Journey breakout session said, it is just about “helping clients do what they want to do.”

Velocity includes preconfigured journeys based on specific client needs, as developed, understood and addressed in all of KPMG previous engagements. Similar to many consultancies, KPMG begins engagements by developing an understanding of clients’ strategic needs and issues, rather than their technology stack. (TBR comment: easy to say, hard to do, especially when a firm has practices built around specific technologies).

Velocity is designed to add value to client engagements (including describing, calculating and being accountable). It will also bring a “tremendous amount of information” and is “highly tailorable,” according to a KPMG leader, who also noted that the platform’s adoption, use and usefulness over time will be key. KPMG leaders said the core aspects of AI — even agentic AI — are all the same, separated only by planning and orchestration. For example, KPMG’s AI Workbench underpins how it is bringing agents and AI-enabled services to its clients and its people. Velocity, then, is a KPMG offering where every step is focused on achieving client outcomes, which comes back to understanding clients’ key business issues, not simply their technology stack.

The launch of Velocity internally (starting in March 2025) into its largest member firms brings to life KPMG’s approach to AI. KPMG expects its member firms to be able to start unlocking the power of Velocity beginning in May, and will launch Velocity externally later in the year. Amid caution on the client side around the adoption and implementation of AI technologies, KPMG’s David Rowlands, head of AI, discussed how KPMG wants to be client zero around AI, helping to ease clients’ ethics and security concerns by working through experimentation and into adoption and scale. Rowlands highlighted the firm’s attention to knowledge and need to fully benefit from AI. Training around AI, including the definition of AI and how to use it; creating trust within AI; and learning effective AI prompts also fit within this strategy, enabling both KPMG and clients to effectively embed AI across people and operational strategies.

 

Velocity, AI and the future of audits

Three other AI-centric comments from KPMG leaders stood out for TBR:

  • With AI, “the road to value is paved with human behavior and change,” according to Rowlands, reflecting the firm’s emphasis on the business over the technology and the importance of change management — a core KPMG consulting strength.
  • Rowlands also noted that AI is a critical national infrastructure, dependent on energy, connectivity and networks, and should be considered a national investment priority and national security issue. In TBR’s view, framing AI this way — not as just a tool or another service to be sold — adds credibility to KPMG’s AI efforts.
  • According to Per Edin, KPMG’s AI leader, “ROI is clear and documented, but still not enough adoption to be as measurable as desired.” In TBR’s view, Edin’s sentiments track closely with TBR’s Voice of the Customer and Voice of the Partner research, which have repeatedly shown that interest and investments in AI have outpaced adoption, particularly at scale.

In a breakout session, KPMG walked through the firm’s well-established KPMG Clara platform, a tool designed to help the firm complete its audits more quickly and accurately. In essence, KPMG creates a digital twin of an organization, reflecting the firm’s understanding of where AI can be applied. KPMG Clara Transaction Scoring enables auditors to deliver what the firm calls “audit evidence” and note “outlier” transactions. According to KPMG leaders, “AI agents perform audit procedures and document results for human review, just like junior staff.”
 
Critically, KPMG Clara audits every transaction, not just a sample of transactions, increasing the likelihood of catching problems, issues and outliers. By flagging high-risk transactions, KPMG can deploy professionals to focus on solving real problems rather than adjudicating false positives or meaningless issues. In TBR’s view, this represents the proverbial “higher-value task” long-promised by robotic process automation, AI and analytics.
 
When pressed by TBR, KPMG leaders said clients were not looking for rate cuts but rather for higher-quality audits and new insights into their operations. Importantly, clients also expect to spend less time on an audit, freeing up professionals’ time: The client can do what they do, and KPMGers can stay focused on reflected issues and generate new insights.
 
TBR remains a bit skeptical, but if clients do not expect a rate cut when KPMG deploys AI to speed up the audit process and instead expect to spend less time internally on what should be a higher-quality audit, TBR considers that a fantastic way to position AI while also reducing KPMG’s professionals’ time. There are two unanswered questions: What happens to the apprenticeship model, in which less-experienced KPMG professionals learn the art, not the science, of audit? And, in a few years, will 95,000 professionals conduct 400,000 audits (twice the current number) or will 50,000 professionals (half the current staff) complete 200,000 audits?
 
Regardless, as the company rolls out internally developed generative AI (GenAI) tools, the learning and experience captured through the firm’s implementation and change management process will undoubtedly be integrated into customer engagements involving third-party solutions. With SAP and Salesforce in attendance, KPMG zeroed in on each vendor’s AI strategy and how the firm plans to support it. To focus on Salesforce, Lohr echoed Salesforce CEO Marc Benioff in calling Agentforce the most successful Salesforce launch ever, which suggests a recognition from KPMG leadership of Salesforce’s agentic AI strategy.
 
For its part, KPMG highlighted the recent launch of an Agentforce Incubator, an experimental experience that can be delivered to clients from any location — a client site, Salesforce event or a KPMG office — to ignite the ideation stage and begin exploring the road map from proof of concept to production. During one-on-one conversations, TBR explored KPMG’s view of its role in the agentic AI, and we found it to be both pragmatic and valuable — similar to how the firm must be opinionated in broader digital transformation engagements.
 
KPMG’s journey to becoming an AI orchestrator will require the firm to take a stance on a vendor-by-vendor basis and arrive on-site with a preconceived understanding of the best path forward for clients given their goals. In addition to having an opinion, KPMG also recognizes it must help facilitate the road maps it lays out to clients, which will involve a heavy change management component, as well as a more technical design and development element. With the Agentforce example, once a targeted business outcome is established, an AI agent needs to be designed and developed to achieve the outcome. In many cases, a customer may lack the internal technical resources necessary to build the agent and tackle the problem. As KPMG avoids vendor agnosticism, the company can focus on building out technical resources with the vendors it chooses, building deeper benches with technical training associated with its strategic partners.

KPMG’s Lakehouse offers unique setting for analyst event

As it did less than 18 months ago when KPMG broke from the traditional analyst event style, the firm did it again. Hosting 62 analysts and dozens of global executives, clients and partners at a flagship Lakehouse facility for two days of both formal and informal interactions, presentations, client use cases and demos, KPMG demonstrated agility in terms of the delivery and engagement format, yet, with a steady hand, continued to execute on its vision with its global solutions — Connected, Powered, Trusted and Elevate — and proven IP, methods and enablers coming together through Velocity.
 
KPMG held one-on-one sessions between analysts and executives midway through the first day so that executives were present and engaged. Additionally, KPMG saved the all-about-AI-and-nothing-else sessions for the second day, which came off as, “We get AI is important, but we are also realistic and keeping our heads on straight and not being ‘me too, me too’ about AI.” KPMG senior executives sat in on both the client case study and platform breakout sessions. Subtle message to analysts: This stuff matters enough across the firm to be worth KPMG partners’ time even if it is not in their area.

Conclusion

As a member of the Big Four, KPMG has brand permission and a breadth of services that are relevant to nearly every role in any enterprise. As the firm executes on its Collective Strategy, TBR believes KPMG will accelerate the scale and completeness of its offerings, building on a solid foundation and expanding the gaps between KPMG and other consulting-led, technology-enabled professional services providers. 

KPMG’s global solutions — Connected, Powered, Trusted and Elevate — which resonate with clients and technology partners, have now been brought together into one transformation framework under KPMG Velocity, providing KPMG’s professionals with clear insight into the firm’s strengths and strategy, and underpinning, in the near future, all KPMG’s transformation engagements. KPMG Velocity’s evolving strategy will challenge KPMG’s leaders to execute on the promise of that transformation during the next wave of macroeconomic pressures, talent management battles and technology revolutions. At the same time, KPMG’s leaders recognize that their priorities are transforming the firm’s go-to-market approach, unlocking the power of the firm’s people, reimagining ways of working, and innovating capabilities and service enhancements. 

Success in executing these priorities, in TBR’s view, will come as KPMG shifts from building a foundation to scaling alongside the growing needs of its clients and as the era of GenAI presents yet another opportunity and challenge. Striking the right balance between elevating the potential of GenAI as a value creator and accounting for commercial and pricing model implications will test the durability of KPMG’s engagement and delivery frameworks. 

Although the firm has placed in motion many of the aforementioned investments over the past 12 to 18 months, the one opportunity that is changing relates to speed. As one enterprise buyer recently explained to TBR: “GenAI will force all services vendors to change. The [ones] who [will] be [the] most successful will be [those] who do it fast.” With speed comes risk — which KPMG fully acknowledges and is why KPMG Velocity’s offering is a differentiator for the firm in the market. With KPMG Velocity, all of KPMG’s multidisciplinary and heritage risk and regulatory considerations have been embedded across each transformation journey to ensure clients can remain compliant and avoid the pitfalls that can often arise during transformation.

Special report contributors: Catie Merrill, Senior Analyst; Kelly Lesizcka, Senior Analyst; Alex Demeule, Analyst; Boz Hristov, Principal Analyst

TBR Announces On-demand Availability of 1Q25 TBR Insights Live Webinars

Technology Business Research, Inc. (TBR) is pleased to announce on-demand availability of all TBR Insights Live webinars that aired during 1Q25. Each session features TBR’s top experts discussing key trends, recent analysis and market dynamics across covered industries.
 
1Q25 webinars can be viewed on our YouTube channel, and presentation decks can be downloaded via the linked titles below.

Explore the full lineup of 1Q25 TBR Insights Live webinars

MWC25: Disruptive Technologies and Business Models Create New Opportunities for the Mobile Ecosystem

  • How the telecom industry intends to derive business outcomes from AI
  • How enterprises are progressing in their digital transformations and incorporating private networks
  • Where in the mobile ecosystem new value is being created and what telcos need to do to generate ROI from new opportunities

Ecosystem Intelligence for IT Services, Cloud and Consultancies: Strategic Insights for 2025 Success

  • How to place strategic ecosystem bets on alliance partners that are well-positioned for the next growth wave
  • How competitors are gaining ground with common alliance partners through sales programs, go-to-market motions and training
  • How to create unique value with alliance partners that resonates with end customers

Digital Transformation Outlook: Strategy Rebound, GenAI Impact and Ecosystems Importance in 2025

  • Why strategy consulting will rebound in 2025, and which consultancies will benefit
  • Discussion of ERP consolidation (e.g., S4Hanna, mainframe modernization)
  • Cloud migration services, custom apps development and workflow management
  • How generative AI (GenAI)-enabled solutions will upend organizational structures and business models for IT services and consultancies, with follow-on effects for partners
  • How the emergence of ecosystem intelligence as a strategic priority will impact IT services companies, consultancies and technology alliance partners

Cloud Market 2025: How GenAI Will Shape the Future

  • How GenAI will offset some of the slowdown in cloud revenue growth
  • How Microsoft will challenge Amazon Web Services’ leadership in IaaS and PaaS
  • How SaaS vendors will monetize GenAI in 2025

6G: How Government Intervention Globally Will Shape the Next Generation of Telecom

  • What spectrum bands 6G will likely leverage
  • How 6G will shape communication service providers’ capex investments
  • How governments might get involved to ensure 6G becomes a reality

AI PCs in 2025: Unlocking Mass Appeal and Overcoming Market Challenges

  • The potential impact of AI PCs on PC refresh cycles in 2025
  • How PC makers will set themselves apart
  • Expectations for the PC market overall in 2025 based on TBR’s latest research and analysis

Navigating GenAI: Insights, Strategies and Opportunities for 2025

  • How GenAI is impacting buyer-vendor relationships and what is next in the evolution of their business models
  • How tech and services companies are using alliances to extend their reach within enterprises and across the larger GenAI — and emerging tech — ecosystem
  • Which consultancies, IT services vendors, cloud and software companies, and infrastructure players are best positioned for the next wave of GenAI adoption

 
Visit the TBR Insights Live page to discover what our experts will be discussing next.

Comcast Business Nears $10B in Annual Revenue and Accelerates Enterprise Growth but Faces Headwinds from Competitive and Macroeconomic Pressures

2025 Comcast Business Analyst Conference, Philadelphia, April 2-3, 2025 — A select group of industry analysts gathered at the Comcast Center in Philadelphia to hear from Comcast Business leaders about the unit’s progress and success with its sales and go-to-market strategies. The central theme of the event was “Everything, Everywhere, All at Once,” which reflects Comcast Business’s ability to provide solutions to its customers through advancements in areas including AI implementation, network technologies, industry partnerships, and acquisitions, including Masergy and Nitel. The event was hosted by CNBC Senior Markets Correspondent Dominic Chu and included a State of the Business session from Comcast Business President Edward Zimmermann, a Strategy & Vision session from Comcast Business Chief Product Officer Bob Victor, and an update on Comcast’s network from Chief Network Officer Elad Nafshi. Also included were panel discussions with senior leadership as well as speaker sessions featuring industry partners Cisco and Intelisys and a Comcast Business customer within the brewing industry.

 

TBR perspective

Since its inception in 2006, Comcast Business has consistently grown into a more formidable competitor in the B2B telecom space. Most of this growth has stemmed from the SMB segment, where Comcast Business’ superior DOCSIS-based, hybrid-fiber coax (HFC) fixed broadband offerings were priced right compared to non-fiber-to-the-premises (FTTP) telco offerings and addressed demand for more bandwidth. Comcast Business’ strategy has evolved over the past decade to target additional growth segments including midmarket businesses and multinational enterprises via the operator’s managed services portfolio, strategic acquisitions including Masergy, and partnerships with global operators spanning 130 countries.
 
Other key growth drivers over the past decade include Comcast Business’ increasing focus on the public sector, including federal agencies, as well as the launch of portfolio segments including Comcast Business Mobile and value-added services in areas including SD-WAN, security and unified communications.
 
The success of Comcast Business’ growth strategies has enabled the company to essentially reach its long-term goal of generating $10 billion in annual revenue (Comcast Business generated $9.7 billion in revenue for full-year 2024). Comcast Business, which increased total revenue by about 5% in 2024, is also outpacing incumbent operators competing in the B2B market in revenue growth as service providers including AT&T, Verizon and Lumen continue to face significant revenue erosion from customers disconnecting from legacy data and voice services.
 
Despite its recent strong momentum, Comcast Business will encounter obstacles as it tries to increase revenue due to headwinds including competition from the expansion of fixed wireless access (FWA) and FTTP services, and macroeconomic pressures that will cause businesses to optimize spending on connectivity services. Despite its ability to sustain revenue growth in 2024, driven by increased revenue from enterprise solutions and higher average rates from small business customers, Comcast Business lost a net of 16,000 customer relationships in 2024 (compared to customer relationship net additions of 17,000 in 2023 and 52,000 in 2022).
 
TBR attributes rising B2B FWA adoption, especially for Verizon and T-Mobile, as the primary driver of the net loss as more businesses are gravitating to FWA for its cost savings over traditional fixed broadband services as well as its greater ease of installation, which is helping to support companies seeking to quickly launch new branch locations. Comcast Business will also face competitive pressures from operators including AT&T, T-Mobile and Verizon that are expanding their FTTP footprints via organic builds and acquisitions, which will give these operators new opportunities to offer converged service plans combining mobile and broadband services. Comcast Business will also continue to face pressures from businesses continuing to migrate off of pay-TV and VoIP (voice over IP) services.
 
We expect that current macroeconomic challenges, including mass layoffs within the private and public sectors and uncertainty around tariff impacts, will create headwinds for Comcast Business and the overall U.S. B2B telecom market. Though network connectivity solutions such as broadband will remain essential to businesses, companies will need to optimize spending to counter macroeconomic pressures.
 
TBR believes these challenges will require Comcast Business to become more dependent on providing a stronger value proposition to retain and grow its customer base in addition to its reliance on the strengths of its network and solutions portfolio. Comcast Business is also addressing these industry shifts by evolving its portfolio of adjacent solutions in areas including secure networking, cybersecurity, and managed IT services to augment revenue from its core broadband services.

Impact and opportunities

Comcast Business is leveraging AI to optimize its network performance and sales and customer service capabilities

Sessions throughout the event discussed how deeper AI implementation will enable Comcast to enhance capabilities such as network performance and sales and customer service, thereby improving overall efficiency and customer experience. For instance, AI integration is enabling Comcast to automate over 99.7% of all software changes that it is making on its network, which is supporting network self-healing capabilities that can quickly resolve outages. These capabilities will help Comcast Business to more effectively retain customers as recent disruptions experienced by rivals, such as AT&T’s two major prolonged network outages in 2024, have resulted in customer losses and tarnished brand images for impacted operators.
 
AI is also enabling Comcast to enhance the cybersecurity of its network, including through the development of a next-generation firewall embedded into the network, which leverages GenAI and does not require dedicated CPE (customer premises equipment).
 
The vendor is also focused on training its customer care and sales teams to more effectively leverage AI to improve customer support and enhance operations. Comcast Business is increasing the number of AIOps use cases and applying AI and machine learning (ML) across its managed solutions platform to improve service delivery, assurance and management, both for customers and the internal teams that support customers (e.g., help desk, network operations center [NOC] and security operations center [SOC]).
 
Comcast expects AI to not only improve network and operational efficiencies but also provide significant revenue-generation opportunities, though the company is still in the early stages of developing strategies to do so. For instance, Comcast’s edge computing resources enable the company to support ultra-low latency speeds of less than 1 millisecond to many of its customers. These capabilities will position Comcast to optimize connectivity and user experience for future advanced AI applications in areas such as AR/VR that will be more dependent on ultra-low latency, though current AI applications such as ChatGPT are not as dependent on ultra-low latency as they are mainly text-based.

Comcast Business continues to accelerate its data speeds to incentivize customers, though industry pricing pressures will hamper connection growth

Throughout the event, Comcast Business promoted its accelerating data speeds, which are aided by network advancements such as DOCSIS 4.0 and mid-split upgrades to Comcast’s HFC network. Enhancements to Comcast Business’ connectivity portfolio include extending the availability of its Dedicated Internet solution and upgrading the service to provide symmetrical download and upload speeds up to 200Mbps over HFC or up to 400Gbps over fiber.
 
Comcast expects to accelerate its Dedicated Internet solution to reach symmetrical speeds of 300Mbps over HFC and reach a total of over three million passings this year. Other updates to the Dedicated Internet solution include adding a network reliability guarantee, which provides SLAs ensuring 99.99% network uptime, and enhanced proactive network monitoring, which enables IT teams to optimize performance.
 
TBR believes these updates will help to attract clients with bandwidth-intensive workloads, especially customers with more stringent SLA requirements necessitating minimal network downtime. However, TBR also recognizes that competitive pricing and the overall value proposition provided by operators are becoming more influential factors in contract wins within the B2B market. As evidenced by the robust uptake of FWA, small businesses are especially concerned with the value they are getting for the price paid, and they are migrating to lower-cost broadband offerings to obtain internet access that more closely meets their needs and aligns with what they are willing to pay.
 
T-Mobile and Verizon are feeding this market shift to “rightsized bandwidth” through clever marketing and customer education about what businesses actually need. Comcast Business will need to demonstrate why its cutting-edge broadband offerings are necessary for its customers in order to justify the premium pricing. It also has an opportunity to further strengthen the value proposition of its value-added services when combined with its core broadband services. Comcast Business Mobile is a key portfolio segment Comcast Business can further leverage to combat pressures from rivals’ FWA services and converged service bundles.
 
Though Comcast Business Mobile connections are not reported by Comcast, TBR believes just a small portion of Comcast Business customers are currently enrolled in the service as only 12% of Comcast’s residential broadband customer were enrolled in Xfinity Mobile in 4Q24. Heavier promotional activity, such as offering free lines for a limited time, could help Comcast Business Mobile compete more aggressively against rival B2B smartphone plans while creating a stickier ecosystem to retain high-value broadband customers long-term. Comcast Business Mobile’s impact is limited to the SMB market, however, as the brand supports a maximum of 20 lines per business customer under Comcast’s current MVNO agreement with Verizon.

Comcast will strengthen its enterprise business and expand sales channels via the Nitel acquisition

The analyst event coincided with Comcast Business closing its acquisition of Chicago-based managed service provider Nitel on April 1 from private equity firm Cinven. Nitel is a NaaS (Network as a Service) provider offering solutions in areas such as networking (including SD-WAN and SASE [Secure Access Service Edge]), cloud services and cybersecurity.
 
Through the purchase, Comcast Business will expand its footprint in the midmarket and enterprise customer segments and gain Nitel’s 6,600 clients across the U.S. within verticals including financial services, healthcare and education. Acquiring Nitel also enables Comcast Business to expand its channel distribution strategy to more effectively target new sales opportunities within the midmarket and enterprise segments.
 
Comcast Business is also gaining AI and software tools from the Nitel acquisition that will enable it to enhance its sales and customer service capabilities. These benefits include robust orchestration capabilities, an instant quoting tool that makes it easier to price and establish deals across multiple vendors and sites, and a digital dashboard that offers a single-pane-of-glass view of deployments.

Conclusion

Comcast Business remains in a relatively strong position within the B2B market as the company continues to outpace its competitors in revenue growth and will continue to expand its client base within the midmarket and large enterprise segments by leveraging its Masergy and Nitel acquisitions. Comcast Business also has an opportunity to increase revenue by refining its international strategy and more deeply leveraging assets — such as its managed services portfolio, partnerships with global operators across 130 countries, and numerous acquisitions including Sky, Masergy, Nitel, Deep Blue Communications and Blueface — that enable it to support multinational corporations.
 
However, SMB, which accounts for the majority of Comcast Business’ revenue, is becoming a more challenging segment in which to grow market share as FWA competition and macroeconomic challenges lead to spending constraints. These headwinds will require Comcast Business to become more focused on enhancing its value proposition to retain and grow its SMB client base and combat competitive pressures in the market.

Sheer Scale of GTC 2025 Reaffirms NVIDIA’s Position at the Epicenter of the AI Revolution

As the undisputed leader of the AI market, NVIDIA and its GPU Technology Conference (GTC) are unmatched compared to other companies and their respective annual events when it comes to the enormous impact they have on the broader information technology market. GTC 2025 took place March 17-21 in San Jose, Calif., with a record-breaking 25,000 in-person attendees — and 300,000 virtual attendees — and nearly 400 exhibitors on-site to showcase solutions built leveraging NVIDIA’s AI and accelerated computing platforms.

NVIDIA GTC 2025: Pioneering the future of AI and accelerated computing

In 2024 NVIDIA CEO and cofounder Jensen Huang called NVIDIA GTC the “Woodstock of AI,” but to lead off the 2025 event’s keynote address at the SAP Center, he aptly changed his phrasing, calling GTC 2025 “the Super Bowl of AI,” adding that “the only difference is that everybody wins at this Super Bowl.”
 
While the degree to which every tech vendor “wins” in AI will vary, NVIDIA currently serves as the rising tide that is lifting all boats — in this case, hardware makers, ISVs, cloud providers, colocation vendors and service providers — to help accelerate market growth despite the economic and geopolitical struggles that have hampered technology spending in the post-COVID era. NVIDIA’s significant investments not as a GPU company but as a platform company — delivering on innovations in full-stack AI and accelerated computing infrastructure and software — have provided much of the foundation upon which vendors across the tech ecosystem continue to build their AI capabilities.
 
During the event, which also took place at the nearby San Jose McEnery Convention Center, Huang shared his vision for the future, emphasizing the immense scale of the inference opportunity while introducing new AI platforms to support what the company sees as the next frontiers of AI. Additionally, he reaffirmed NVIDIA’s commitment to supporting the entire AI ecosystem by building AI platforms, rather than AI solutions, to drive coinnovation and create value across the entire ecosystem.

The transformation of traditional data centers into AI factories represents a $1 trillion opportunity

The introduction of ChatGPT in November 2022 captured the attention of businesses around the world and marked the beginning of the generative AI (GenAI) revolution. Since then, organizations across all industries have invested in the exploration of GenAI technology and are increasingly transitioning from the prototyping phase to the deployment phase, leveraging the power of inference to create intelligent agents, power autonomous vehicles and drive other operational efficiencies. As AI innovation persists, driven largely by the vision of Huang and the increasingly capital-rich company behind him, new AI paradigms are emerging and NVIDIA is helping the entire AI ecosystem to prepare and adapt.

The rise of reasoning

On Jan. 27, otherwise known as DeepSeek Monday, NVIDIA stock closed the day down 17.0% from the previous day’s trading session, with investors believing DeepSeek’s innovations would materially reduce the total addressable market for AI infrastructure. DeepSeek claimed that by using a combination of model compression and other software optimization techniques, it had vastly reduced the amount of time and resources required to train its competitive AI reasoning model, DeepSeek-R1. However, at GTC 2025, NVIDIA argued that investors misunderstood implications on the inference side of the AI model equation.
 
Traditional knowledge-based models can quickly return answers to users’ queries, but because basic knowledge-based models rely solely on the corpus of data that they are trained on, they are limited in their ability to address more complex AI use cases. To enhance the quality of model outputs, AI model developers are increasingly leveraging post-training techniques such as fine-tuning, reinforcement learning, distillation, search methods and best-of-n sampling. However, more recently test-time scaling, also known as long thinking, has emerged as a technique to vastly expand the reasoning capabilities of AI models, allowing them to address increasingly complex queries and use cases.

From one scaling law to three

In the past, pre-training scaling was the single law dictating how applying compute resources would impact model performance, with model performance improving as pre-training compute resources increased. However, at GTC 2025, NVIDIA explained two additional scaling laws in effect — post-training scaling and test-time scaling. As their names suggest, model pre-training and post-training are on the AI model training side of the equation. However, test-time scaling takes place during inference, allocating more computational resources during the inference phase to allow a model to reason through several potential responses before outputting the best answer.
 
Traditional AI models operate quickly, generating hundreds of tokens to output a response. However, with test-time scaling, reasoning models generate thousands or even tens of thousands of thinking tokens before outputting an answer. As such, NVIDIA expects the new world of AI reasoning to drive more than 100 times the token generation, equating to more than 100 times the revenue opportunity for AI factories.
 
During an exclusive session with industry analysts, Huang said, “Inference is the hardest computing at scale problem [the world has ever seen],” dispelling the misnomer that inference is somehow easier and demands fewer resources than training while also indirectly supporting Huang’s belief that the transformation of traditional data centers into AI factories will drive total data center capital expenditures (capex) to $1 trillion or more by 2028.
 

Graph: NVIDIA Revenue, Growth and Projections (Source: TBR)

NVIDIA Revenue, Growth and Projections (Source: TBR) — If you believe you have access to TBR’s NVIDIA research via your employer’s enterprise license or would like to learn how to access the full research, click here.


 
While on the surface, $1 trillion in data center capex by 2028 sounds like a lofty threshold, TBR believes the capex amount and timeline are feasible considering NVIDIA’s estimate that 2024 data center capex was around $400 billion.
 
Additionally, during 1Q25, announcements centered on investment commitments to build out data centers have become increasingly common, and TBR expects this trend to only accelerate over the next few years. For example, in January the Trump administration announced the Stargate Project with the intent to invest $500 billion over the next four years to build new AI infrastructure in the United States.
 
However, it is worth noting that Stargate’s $500 billion figure represents more than just AI servers; it includes other items such as the construction of new energy infrastructure to power data centers. TBR believes the same holds true for NVIDIA’s $1 trillion figure, especially when considering TBR’s 2024 total AI server market estimate of $39 billion.

The more you buy, the more you make: NVIDIA innovates to maximize potential AI factory revenue

To support the burgeoning demands of AI, NVIDIA is staying true to the playbook through which it has already derived so much success — investing in platform innovation and the support of its growing partner ecosystem to drive the adoption of AI technology across all industries.

AI factory revenue relies on user productivity

Reasoning capabilities allow models to meet the demands of a wider range of increasingly complex AI use cases. Although the revenue opportunity of AI factories increases as AI reasoning drives an exponential rise in token generation, expanding token generation also creates bottlenecks within AI factories and inevitably there is a tradeoff. To maximize revenue potential, AI factories must optimize the balance between token volume and cost per token.
 
From the perspective of an AI inference service user, experience comes down to the speed at which answers are generated and the accuracy of those answers. Accuracy is tied directly to the underlying AI model(s) powering the service and can be thought of as a constant variable in this scenario, while the speed at which answers are generated for a single user is dictated by the rate of output token generation for that specific user. Having more GPUs dedicated to serving a single user results in an increased rate of output token generation for that user and is something that users are typically willing to pay a premium for.
 
However, in general, as more GPUs are dedicated to serving a single user, the overall output token generation of the AI factory falls. On the opposite end of the spectrum, an AI factory can maximize its overall output token generation by changing GPU resource allocations to serve a greater number of users at the same time; however, this has a negative impact on the rate of output tokens generated per user, increasing request latency and thereby detracting from the user’s experience.
 
As NVIDIA noted during the event, to maximize revenue, AI factories must optimize the balance of total factory output token generation and the rate of output token generation per user. However, once the optimal allocation of GPU resources is determined, revenue opportunity hits a threshold. As such, to increase the productivity and revenue opportunity of AI factories, NVIDIA supports the AI ecosystem with its investments in the development of increasingly performant GPUs, allowing for greater total factory output token generation as well as increased rates of output token generation per user.
 
During his keynote address, Huang laid out NVIDIA’s four-year GPU road map, detailing the upcoming Blackwell Ultra as well as the NVIDIA GB300 NVL72 rack, which leverages Blackwell Ultra and features an updated NVL72 design for improved energy efficiency and serviceability. Additionally, he discussed the company’s Vera Rubin architecture, which is set for release in late 2026 and marks the shift from HBM3/HBM3e to HBM4 memory, as well as Vera Rubin Ultra, which is expected in 2027 and will leverage HBM4e memory to deliver higher memory bandwidth. To round out NVIDIA’s four-year road map, Huang announced the company’s Feynman GPU architecture, which is slated for release in 2028.

Scale up before you scale out, but NVIDIA supports both

In combination with NVIDIA’s updated GPU architecture road map, Huang revealed preliminary technical specifications for the Vera Rubin NVL144 and Rubin Ultra NVL576 racks, with each system being built on iterative generations of the company’s ConnectX SuperNIC and NVLink technologies, promising stronger networking performance with respect to increased bandwidth and higher throughput. NVIDIA’s growing focus on NVL rack systems underscores Huang’s philosophy that organizations should scale up before they scale out, prioritizing the deployment of fewer densely configured AI systems compared to a greater number of less powerful systems to drive simplicity and workload efficiency.
 

Graph: 2024 Data Center GPU Market Share (Source: TBR)

2024 Data Center GPU Market Share (Source: TBR) — If you believe you have access to TBR’s NVIDIA research via your employer’s enterprise license or would like to learn how to access the full research, click here.


 
Networking has and continues to become more integral to NVIDIA’s business as the company’s industry-leading advancements in accelerated compute have necessitated full-stack AI infrastructure innovation. While NVIDIA drives accelerated computing efficiency on and close to the motherboard through the design of increasingly high-performance GPUs and CPUs and its ongoing investments in ConnectX and NVLink, the company is also heavily invested in driving AI infrastructure efficiency through its networking platform investments in Quantum-X InfiniBand and Spectrum-X Ethernet.
 
Although copper is well suited for short-distance data transmissions, fiber optics is more effective over long distances. As such, the scale-out of AI factories requires an incredible number of optical transceivers to connect every NIC (network interface card) to every switch, representing the single largest hardware component in a typical AI data center. NVIDIA estimates that optical transceivers consume approximately 10% of total computing power in most AI data centers. During his keynote address, Huang announced NVIDIA Photonics — what the company describes as a coinvention across an ecosystem of copacked optics partners — to reduce power consumption and the number of discrete components in an AI data center.
 
Leveraging components from partners, including TSMC, Sumitomo and Corning, NVIDIA Photonics allows NVIDIA to replace pluggable optical transceivers with optical engines that are copackaged with the switch ASIC. This allows optical fibers to plug directly into the switch with the onboard optical engine processing and converting incoming data — in the form of optical signals — into electrical signals that can then be immediately processed by the switch. Liquid-cooled Quantum-X Photonic switch systems are expected to become available later this year ahead of the Spectrum-X Photonic switch systems that are coming in 2026. NVIDIA claims that the new systems improve power efficiency by 3.5x while also delivering 10x higher resiliency and 1.3x faster time to deploy compared to traditional AI data center architectures leveraging pluggable optical transceivers.

Securing the developer base

Adjacent to what the company is doing in the data center, NVIDIA announced other, more accessible Blackwell-based hardware platforms, including RTX PRO Series GPUs, DGX Spark and DGX Station, at GTC 2025. At CES (Consumer Electronics Show) 2025 in January, NVIDIA made two major announcements: Project DIGITS, a personal AI supercomputer that provides AI researchers, data scientists and students with access to the Grace Blackwell platform; and the next-generation GeForce RTX 50 Series of consumer desktop and laptop GPUs for gamers, creators and developers.
 
Building on these announcements, at GTC 2025 NVIDIA introduced DGX Spark, the new name of the previously announced Project DIGITS, leveraging NVIDIA GB10 Grace Blackwell Superchip and ConnectX-7 to deliver 1,000 AI TFLOPS (tera floating-point operations per second) performance in an energy-efficient and compact form factor. DGX Spark will come pre-installed with the NVIDIA AI software stack to support local prototyping, fine-tuning and inferencing of models with up to 200 billion parameters, and NVIDIA OEM partners ASUS, Dell Technologies, HP Inc. and Lenovo are already building their own branded versions.
 
To complement its recently unveiled GeForce RTX 50 Series, NVIDIA announced a comprehensive lineup of RTX PRO Series GPUs for laptops, desktops and servers with “PRO” denoting the solutions’ intent to support enterprise applications. At the top end of the lineup, RTX PRO 6000 will deliver up to 4,000 AI TFLOPS performance, making it the most powerful discrete desktop GPU ever created. While DGX Spark systems will be available beginning in July, DGX Station is expected to be released toward the end of the year. DGX Station promises to be the highest-performing desktop AI supercomputer, featuring the GB300 Grace Blackwell Ultra Desktop Superchip and ConnectX-8, with OEM partners, including ASUS, Box, Dell Technologies, HP Inc., Lambda and Supermicro, building systems. Together, these announcements highlight NVIDIA’s commitment to democratizing AI and supporting developers.

Software is the most important feature of NVIDIA GPUs

In TBR’s 1Q24 Semiconductor Market Landscape, NVIDIA led all vendors in terms of trailing-12 month (TTM) corporate revenue growth, with hardware revenue accounting for an estimated 88.9% of the company’s TTM top line. However, while NVIDIA’s industry-leading top-line growth continues to be driven primarily by increasing GPU and AI infrastructure systems sales, the reason customers choose NVIDIA hardware ultimately boils down to two interrelated factors: the company’s developer ecosystem, and its AI platform strategy.

The CUDA advantage

In 2006 NVIDIA introduced CUDA (Compute Unified Device Architecture), a coding language and framework purpose-built to enable the acceleration of workloads beyond graphics. With CUDA, developers gained the ability to code applications optimized to run on NVIDIA GPUs. Since CUDA’s inception, NVIDIA has relentlessly invested in strengthening CUDA, supporting backward compatibility, publishing new CUDA libraries, and giving developers new resources to optimize the performance and simplify the building of applications.
 
As such, many legacy AI applications and libraries are rooted in CUDA, whose documentation is light years ahead of competing platforms, such as AMD ROCm. With respect to driving AI efficiency, several NVIDIA executives and spokespeople at GTC 2025 circled back to the notion that, when it comes to enabling the most complex AI workloads of today and tomorrow, software optimization is as important as, if not more important than, infrastructure innovation and optimization, underscoring the unique value behind NVIDIA’s CUDA-optimized GPUs. In short, at the heart of NVIDIA’s comprehensive AI stack and competitive advantage is CUDA, and as Huang emphasized to the attending industry analysts, “Software is the most important feature of NVIDIA GPUs.”

A new framework for AI inference

As the AI inference boom materializes, NVIDIA has leveraged the programmability of its GPUs to optimize the performance of reasoning models at scale, with Huang introducing NVIDIA Dynamo at GTC 2025. Dynamo is an open-source modular inference framework that was designed to serve GenAI models in multinode distributed environments and specifically developed for accelerating and scaling AI reasoning models to maximize token revenue generation.
 
The framework leverages a technique called “disaggregated serving,” which separates the processing of input tokens in the prefill phase of inference from the processing of output tokens in the decode phase. Traditional large language model (LLM) deployments leverage a single GPU or GPU node for both the prefill and decode phases, but each phase has different resource requirements, with prefill being compute-bound and decode being memory-bound. As NVIDIA’s VP of Accelerated Computing Ian Buck put it, “Dynamo is the Kubernetes of GPU orchestration.”
 
To optimize the utilization of GPU resources for distributed inference, Dynamo’s Planner feature continuously monitors GPU capacity metrics in distributed inference environments to make real-time decisions on whether to serve incoming user requests using disaggregated or aggregated serving while also selecting and dynamically shifting GPU resources to serve prefill or decode inference phases.
 
To further drive inference efficiencies by reducing request latency and time to first token, Dynamo has a Smart Router feature to minimize key value (KV) cache re-computation. KV cache can be thought of as the model’s contextual understanding of a user’s input. As the size of the input increases, KV cache computation increases quadratically, and if the same request is frequently executed, this can lead to excessive KV cache re-computation, reducing inference efficiency. Dynamo Smart Router works by assigning an overlap score to each new inference request as it arrives and then using that overlap score to intelligently route the request to the best-suited resource — i.e., whichever available resource has the highest overlap score between its KV cache and the user’s request — minimizing KV cache recomputation and freeing up GPU resources.
 
Additionally, Dynamo leans on its Distributed KV Cache Manager feature to support both distributed and disaggregated inference serving and to offer hierarchical caching capabilities. Calculating KV cache is resource intensive, but as AI demand increases, so does the volume of KV cache that must be stored to minimize KV cache recomputation. Dynamo Distributed KV Cache Manager leverages advanced caching policies to prioritize the placement of frequently accessed data closer to the GPU, with less accessed data being offloaded farther from the GPU.
 
As such, the hottest KV cache data is stored on GPU memory with progressively colder data being offloaded to shared CPU host memory, solid-state drives (SSDs) or networked object storage. Leveraging these key features, NVIDIA claims Dynamo maximizes resource utilization, yielding up to 30 times higher performance for AI factories running reasoning models like DeepSeek-R1 on NVIDIA Blackwell. Additionally, NVIDIA leaders state that while designed specifically for the inference of AI reasoning models, Dynamo can double token generation when applied to traditional knowledge-based LLMs on NVIDIA Hopper.
 

The Super Bowl but everybody wins

NVIDIA’s astronomical revenue growth and relentless innovation road map aside, perhaps nothing emphasizes the degree of importance the company holds over the future of the entire AI market more than the number of partners that are clamoring to gain a foothold using NVIDIA as a launching point. The San Jose McEnery Convention Center was filled with nearly 400 exhibitors showcasing how NVIDIA’s AI and accelerated computing platforms are driving innovation across all industries. NVIDIA GTC is no longer a conference highlighting the innovations of a single company; it is the epicenter of showcasing AI opportunity, and every company that wishes to play a role in the market was in attendance.
 
The broad swath of NVIDIA’s partner ecosystem was represented. Infrastructure OEMs and ODMs displayed systems built on NVIDIA reference architectures, while NVIDIA inception startups highlighted their own diverse codeveloped AI solutions. However, perhaps the most compelling and largest-scale example of NVIDIA relying on its partners to deliver AI solutions to end customers came from the company’s global systems integrator (GSI) partners.

NVIDIA provides the platform; partners provide the solution

The world’s leading GSIs, including Accenture, Deloitte, EY, Infosys and Tata Consultancy Services (TCS), all showcased how they are leveraging NVIDIA’s AI Enterprise software platform — comprising NIMs, NeMo and Blueprints — to help customers build and deploy their own customized AI solutions with a heavy emphasis on agentic AI. While some of the largest enterprises in the world have the talent required to build bespoke AI solutions, many other organizations rely on NVIDIA-certified GSI partners with training and expertise in NVIDIA’s AI technologies to develop and deploy AI solutions.
 
Agentic AI has emerged as the next frontier of AI, using reasoning and iterative planning to solve complex, multistep problems autonomously, leading to enhanced productivity and user experiences. NVIDIA AI Enterprise’s tools help make this possible, and at GTC 2025, NVIDIA business leaders shed light on three overarching reasons why NVIDIA AI Enterprise has resonated with end customers and NVIDIA partners alike.
 
First, NVIDIA AI Enterprise builds on CUDA to deliver software-optimized full-stack acceleration, much like other NVIDIA AI platforms. Business leaders essentially explained NIMs — the building blocks of AI Enterprise — as an opinionated way of running a GenAI model on a GPU in the most efficient way possible.
 
Second, NVIDIA AI Enterprise is enterprise grade, meaning that the thousands of first- and third-party libraries constituting the platform are constantly maintained with AI copilots scanning for security threats and AI agents patching software autonomously. Additionally, enterprises demand commitments to maintenance and standard APIs that are not going to change, and NVIDIA AI Enterprise ticks these boxes while also offering tiered levels of support services on top of the platform.
 
Finally, because NIMs are containerized, based on Kubernetes, AI Enterprise is extremely portable, allowing the platform to deliver a consistent experience across a variety of environments.

Autonomous vehicles are the tip of the physical AI iceberg

Several of NVIDIA’s automotive partners also attended GTC 2025, displaying their vehicles inside and outside the convention center. These partners all leverage at least one of NVIDIA’s three computing platforms comprising the company’s end-to-end solutions for autonomous vehicles, with several partners leveraging NVIDIA’s entire platform — including General Motors (GM), whose adoption of NVIDIA AI, simulation and accelerated compute was announced by Huang during the GTC 2025 keynote address.
 
While autonomous vehicles are perhaps the most tangible example, NVIDIA’s three computer systems can be used to build robots of all kinds, ranging from industrial robots used on manufacturing lines to surgical robots supporting the healthcare industry. The three computers required to build physical AI include NVIDIA DGX, which is leveraged for model pre-training and post-training; NVIDIA OVX, which is leveraged for simulation to further train, test and validate physical AI models; and NVIDIA AGX, which acts as the robot runtime and is used to safely deploy distilled physical AI models in the real world.
 
Following the emergence of agentic AI, NVIDIA sees physical AI as the next wave of artificial intelligence, and the company has already codeveloped foundation models and simulation frameworks to support advancements in the field with industry-leading partners, such as Disney Research and Google DeepMind.

Conclusion

The sheer scale of NVIDIA GTC 2025 reaffirmed NVIDIA’s position at the epicenter of the AI revolution, with Huang’s keynote address filling all the available seating in the SAP Center. Born from Huang’s long-standing vision of accelerating workloads by applying parallel processing, NVIDIA’s relentless investments in the R&D of the entire AI stack — from GPUs to interconnect and software platforms to developer resources — remains the driving force behind the AI giant’s success and seemingly insurmountable lead over competitors.
 
NVIDIA’s first-mover advantage in accelerated computing was predicated on the company’s CUDA platform and its ability to allow developers to optimize applications running on NVIDIA GPUs. Nearly 20 years later, NVIDIA continues to leverage CUDA and its robust ecosystem of developers to create innovative AI platforms, such as Omniverse and AI Enterprise, that attract partners from every corner of the technology ecosystem. By swimming in its own lane and relying on its growing NVIDIA Partner Network to deliver AI systems and solutions to end customers, NVIDIA has built an unrivaled ecosystem of partners whose actions on the front lines with end customers facilitate the near-infinite gravity behind the company’s AI platforms.