Larry Dignan

Editor in Chief of Constellation Insights Constellation Research

About Larry Dignan: Dignan was most recently Celonis Media’s Editor-in-Chief, where he sat at the intersection of media and marketing. He is the former Editor-in-Chief of ZDNet and has covered the technology industry and transformation trends for more than two decades, publishing articles in CNET, Knowledge@Wharton, Wall Street Week, Interactive Week, The New York Times, and Financial Planning. He is also an Adjunct Professor at Temple University and a member of the Advisory Board for The Fox Business School's Institute of Business and Information Technology.&nbsp; <br>Constellation Insights does the following: Cover the buy side and sell side of enterprise tech with news, analysis, profiles, interviews, and event coverage of vendors, as well as Constellation Research's community and… Read more

CxOs are cautiously optimistic about their AI and automation efforts, but do need to move beyond proofs of concepts into production.

That's the backdrop to Constellation Research's 2025 CxO Confidence Survey and the firm's 2025 AI Survey.

The short version according to Constellation Research CEO R "Ray" Wang is that CxOs are looking to innovation, automation and AI for competitive advantage.

• CCE 2025: AI agents: Dreams, reality and what's next

Here's a look at some of the takeaways from our twin surveys.

• 52% of CxOs see an improved business climate with 49% remaining cautious.
• This split remains for IT budgets. Nearly half of CxOs see IT budgets increasing.
• 52% of budgets are dedicated to strategic differentiation with 33% focused on revenue and growth and 27% on operational efficiency.
• 61% of CxOs are planning proofs of concepts in AI with automation second at 46% and analytics at 33%.
• Hyperscalers are seen as the go-to co-innovation providers with AWS, Microsoft and Google Cloud ranking as the top three vendors among CxOs.
• On the AI front, 60% of CxOs responding to the Constellation Research 2025 AI Survey said they're primarily focused on working with vendors or systems integrators to build semi-custom applications.
• Half of the AI survey respondents said they are building homegrown applications based on cloud-based machine learning and AI services.
• Half of CxOs are seeing modest returns on AI efforts. One in five achieved returns that meet or exceed initial investments.

Larry Dignan

Editor in Chief of Constellation Insights Constellation Research

About Larry Dignan: Dignan was most recently Celonis Media’s Editor-in-Chief, where he sat at the intersection of media and marketing. He is the former Editor-in-Chief of ZDNet and has covered the technology industry and transformation trends for more than two decades, publishing articles in CNET, Knowledge@Wharton, Wall Street Week, Interactive Week, The New York Times, and Financial Planning. He is also an Adjunct Professor at Temple University and a member of the Advisory Board for The Fox Business School's Institute of Business and Information Technology.&nbsp; <br>Constellation Insights does the following: Cover the buy side and sell side of enterprise tech with news, analysis, profiles, interviews, and event coverage of vendors, as well as Constellation Research's community and… Read more

SavATree, a landscape services company that operates across 40 US states, generated $1.5 million in returns by better engaging and training its frontline workers.

The company was a Supernova 2025 award winner and highlights how engaging frontline workers is increasingly critical to success. SavATree, which deployed UKG Pro Learning, detailed its project at Constellation Research's Connected Enterprise conference.

Sarah Cusack explains the SavATree project at CCE 2025.

I caught up with Sarah Cusack, director of learning and development at SavATree to talk about the project and lessons learned. Here's a look at the takeaways.

Career advancement for field staff needs to be easily accessible. “This project was all about taking what was previously a very basic pen and paper system for training modules, and putting that into a UKG learning management system,” said Cusack, who noted that the previous system hurt engagement due to the manual and slow process.

To make training accessible, the primary delivery method for frontline workers is a mobile device. “We made training more accessible to our field employees through their mobile devices, making it available in the languages that they all speak, and reducing our turnover that was related to career development and promotions,” said Cusack.

The system was a win-win for the company, which benefited from lower turnover, and the employees, who advanced careers and pay quickly.

Delivering returns. Cusak said training for employees needs to drive returns. The primary KPI was a decrease in turnover. Better engagement also translates into better service.

“We saw a 6% decrease in turnover before this program and after that was directly tied to the lead reason of promotion or career advancement,” said Cusak. “Overall, we had about $1.5 million in cost savings, taking into account the reduced turnover, increased engagement, decreased time to fill positions.”

The breakdown went like this:

• Decreased turnover ($255,840 savings): In the two months before the initial rollout of the SDAP program in UKG Pro Learning, 25% of hourly employees cited “career development/promotion” as their primary reason for leaving during exit interviews. In the two months after rollout, less than 10% of hourly employees in that same service line cited “career development/promotion” as their primary reason for leaving.
• Increased employee engagement ($636,480 savings): Leveraging its annual employee engagement survey, SavATree determined that disengaged employees are 18% less productive than engaged employees. Two years after implementing the SDAP program in UKG, disengagement in the main service line fell to 50%, resulting in significant productivity gains.
• Decreased recruiting costs ($392,000 savings): Programmatic recruiting spend, which has the most reliable cost data.
• Decreased training costs ($254,000 savings): Training time and materials.
• Decreased print material costs ($15,000 savings): Reduction in paper usage and printing costs.

Gamification can drive employee engagement. Cusak said a leaderboard where everyone in the company could compare training and learning relative to peers drove engagement.

“The leader boards and competition always sparks a little bit more interest in training,” said Cusak.

• Frontline workers get their AI moment
• Augmenting and Accelerating Frontline Productivity

Balancing AI adoption with authenticity. Cusak said there’s a line enterprises have to walk between using AI and remaining authentic to the frontline worker. “When it comes to things that are really important to our people, we're keeping those as authentic as possible,” said Cusak, who said AI can be used for some content creation, but the things that directly reach frontline workers need a human touch.

Cusak said:

“What we're doing with our approach is making sure we're keeping what is at the heart authentic. Messages from our leadership and frontline leaders will still be authentic. When it's something concerning our mission, vision or values, those things are always authentic and 100% human.”

Larry Dignan

Editor in Chief of Constellation Insights Constellation Research

About Larry Dignan: Dignan was most recently Celonis Media’s Editor-in-Chief, where he sat at the intersection of media and marketing. He is the former Editor-in-Chief of ZDNet and has covered the technology industry and transformation trends for more than two decades, publishing articles in CNET, Knowledge@Wharton, Wall Street Week, Interactive Week, The New York Times, and Financial Planning. He is also an Adjunct Professor at Temple University and a member of the Advisory Board for The Fox Business School's Institute of Business and Information Technology.&nbsp; <br>Constellation Insights does the following: Cover the buy side and sell side of enterprise tech with news, analysis, profiles, interviews, and event coverage of vendors, as well as Constellation Research's community and… Read more

Workday is rounding out its AI strategy, building out its platform with multiple tuck-in acquisitions and looking to become an AI agent player because it can leverage its unified HR and finance data.

The company's third quarter earnings were better than expected, but also highlight how the grand AI strategy, which was outlined at Workday Rising, is just starting.

Workday reported third quarter net income of $252 million, or 94 cents a share, on revenue of $2.43 billion, up 12.6% from a year ago. Non-GAAP earnings were $2.32 a share, 15 cents ahead of Wall Street estimates. As for the outlook, Workday projected fourth quarter subscription revenue growth of 15.5%. For fiscal 2026, Workday projected subscription revenue growth of 14.4%. Fourth quarter operating margin guidance of 28.5% was slightly lower than expected. Workday has had a busy few months.

• Workday acquires Pipedream, launches midmarket focused Workday GO
• Workday launches Workday Data Cloud, Workday Build, more AI agents
• Workday spends $1.1 billion on Sana, aims to be front door to work
• Workday buys Paradox to build out its recruiting suite, delivers strong Q2

"I've been on the road a lot lately, meeting with our customers and prospects, and they're all saying the same thing. They see the potential of AI but they're stuck with disconnected systems, bad data and closed platforms. That's where Workday gives them the ultimate advantage by unifying HR and finance on one intelligent platform," said Workday CEO Carl Eschenbach.

The catch is that Workday has just closed key acquisitions to build out its intelligent platform and announced a few more designed to connect AI agents broadly across the enterprise.

"While other vendors confuse the market with thousands of overlapping general purpose agents, we're focused on what we do best, and that is building powerful agents for HR and finance that deliver real ROI and measurable business value," said Eschenbach, who said data quality is hindering enterprise AI.

If this theme sounds familiar it's because you've heard something similar from multiple enterprise software vendors. You've also heard the same argument from all the AI-driven startups looking to become the new SaaS leaders.

Related research:

• Constellation Research's CXO Confidence Survey
• Constellation Research's 2025 AI Survey
• Decision Velocity in the Agentic Era: Architecting for Decision Automation
• Augmenting and Accelerating Frontline Productivity
• Workday Extend Writes Its Next Chapter: AI
• Case Study - Workday Austrial Payroll

Making sense of Workday's acquisitions

Workday has been on a run of tuck-in acquisitions since its February 2024 purchase of HiredScore. The company just announced the acquisition of Pipedream and closed Sana, which is billed as the future front door to Workday. Paradox, Flowise and Evisort are all deals that are designed to expand Workday's AI agent ambitions.

• Workday acquires Flowise to speed up building AI agents, automated workflows
• Workday aims to be system of record for AI agents, digital labor

These tuck-in deals have largely created Workday's AI agent flywheel. These acquisitions also give Workday attachments to sell to core platforms. For instance, Eschenbach said on Workday's third quarter earnings call that the company is selling Paradox, which focuses on frontline workers, attached to its recruiting software. HiredScore also rides along with recruiting.

Eschenbach said:

"We have the industry-leading AI recruiting platform out there today. At the same time, this is now a new product that is a land-only product for our sales force who can now go and sell Paradox not only on top of Workday or back into our installed base, but also into our competitors' environment. In fact, a significant portion of their existing customers aren't Workday today. And we're going to continue to leverage that go-to-market model, so it gives us another land product without someone having to decide completely on Workday, HR or finance, they can go just with Paradox. And I've seen that come up multiple times just in the first 60 days of us having this great asset."

As for the returns, Eschenbach noted that "for every dollar of recruiting we sell, we sell about $2.50 of HiredScore on top of it." Evisort, a document intelligence for contract management company acquired in Sept. 2024, is a business growing at a triple digit clip for Workday. And Paradox opens the frontline worker recruiting market to Workday.

Zane Rowe, Workday CFO, said both Sana and Paradox are contributing 1.5 points to the company's fourth quarter subscription revenue outlook.

Sana will follow a similar playbook and be sold on top of Workday Learning.

"And then obviously, we're going to refresh our UI/UX, leveraging the Sana platform going forward," said Eschenbach.

Gerrit Kazmaier, President of Product and Technology, said on the third quarter earnings call that Sana will be the "leading UI experience for Workday" and a "complete conversational experience."

He said:

"Imagine every employee having access to HR and finance AI at scale. What that means in cost reduction. On the other side, you can see what drives that interest. And thirdly, Sana goes much more beyond that. And I would recommend you look at the big picture with also Pipedream, adding 3,000 connectors to the Sana platform, which now allows our customers to take Sana knowledge management actions in Workday and the actions that Pipedream adds to really drive enterprise-wide AI transformation with that model."

What Workday is doing is acquiring add-on features and front-ends to the company's HR and financial data.

Using a multi-cloud approach built on AWS and Google Cloud, Workday is mandating that customers move from its own data centers to public cloud. The thinking is that Workday will be able to leverage best-of-breed tooling and spin out innovation faster.

Workday customers will have to update Workday tenant URLs and reconfigure integrations.

The big picture

What Workday is really working toward is an army of controlled AI agents focused on driving enterprise productivity and processes. But to do that you need the data and process intelligence.

Eschenbach said during Workday's Analyst/Investor Day at Workday Rising: "It's not about the quantity of agents you're bringing to market. It's the quality of agents and they have to drive real business value. They have to drive real outcomes."

Kazmaier said AI will be the new UI and enterprise vendors will have to provide leading experiences or lose share.

Here's what Kazmaier outlined as the ingredients to deliver on Workday's ambition. Speaking on Workday's earnings call, Kazmaier outlined the key ingredients to deploying agentic AI:

1. "The first thing that you need is a vast set of data, which basically describes the domain, the domain of finance, the domain of HR. And you need a vast set of data that basically codifies how their data is being used."
2. "You need to have strong semantics and clarity about what data we present. You need to have a data model that defines what data element represents what entity in the business? What do they relate on? And what are the rules for these business entities? They have integrity, they have meaning, they have purpose."
3. "You need to have a business process system, which now basically tells you how to activate this data in a way that can drive towards a business outcome. Even more so, you need to have clarity on what that business outcome is."

Workday's argument is that its consolidated data set and process data is the differentiator. The bet is that data and process drive AI not the other way around.

Larry Dignan

Editor in Chief of Constellation Insights Constellation Research

About Larry Dignan: Dignan was most recently Celonis Media’s Editor-in-Chief, where he sat at the intersection of media and marketing. He is the former Editor-in-Chief of ZDNet and has covered the technology industry and transformation trends for more than two decades, publishing articles in CNET, Knowledge@Wharton, Wall Street Week, Interactive Week, The New York Times, and Financial Planning. He is also an Adjunct Professor at Temple University and a member of the Advisory Board for The Fox Business School's Institute of Business and Information Technology.&nbsp; <br>Constellation Insights does the following: Cover the buy side and sell side of enterprise tech with news, analysis, profiles, interviews, and event coverage of vendors, as well as Constellation Research's community and… Read more

Dell Technologies saw record AI server orders in the third quarter and raised its fiscal 2026 AI shipment guidance to $25 billion, up 150% from a year ago.

The company reported third quarter earnings of $2.28 a share on revenue of $27 billion, up 11% from a year ago. Non-GAAP earnings in the third quarter were $2.59 a share. Wall Street was expecting Dell to report non-GAAP earnings of $2.48 a share on revenue of $27.3 billion.

Dell also named David Kennedy as CFO on a permanent basis. He was interim CFO. Kennedy said fiscal 2026 revenue will be $111.7 billion for the year.

Jeff Clarke, chief operating officer of Dell, said the company has landed $30 billion in AI server orders year to date. "Our five-quarter pipeline is multiples of our $18.4 billion backlog with a mix of neocloud, sovereign and enterprise customers," said Clarke, who noted that Dell is building high-performance systems as well as complex clusters.

• Dell brings more automation to Nvidia AI factory deployments

Like recent quarters, Dell's growth was powered by its infrastructure solutions group. The client solutions group has struggled to deliver revenue growth.

For the infrastructure unit, Dell reported third quarter revenue of $14.1 billion, up 24% from a year ago. Operating income was $1.7 billion, up 16% from a year ago. Servers and networking revenue was $10.2 billion, up 37% from a year ago, and storage revenue fell 1% to $4 billion.

For the PC unit, Dell reported operating income of $748 million in the third quarter on revenue of $12.5 billion, up 3% from a year ago. Commercial client revenue was up 5% and consumer revenue fell 7%.

As for the outlook, Dell projected the following:

• Fourth quarter revenue will be between $31 billion and $3 billion, up 32% from a year ago. Fourth quarter non-GAAP earnings will be $3.50 a share.
• Fiscal 2026 AI server shipments will be about $25 billion, up 150%.
• Fiscal 2026 non-GAAP earnings will be $9.92 a share on revenue of $111.2 billion and $111.2 billion, up 17%.

On the earnings call, Clarke said:

• Dell has AI racks operational within 24 to 36 hours of delivery with uptime topping 99%.
• The company shipped $5.6 billion in AI servers in the quarter. 
• Traditional server demand grew double-digits in EMEA and North America growth accelerating. 
• All-flash array storage systems had double digit demand growth. 
• On the supply chain Clarke said: "We are well positioned across our commodity basket - Q3 was deflationary, and our outlook for Q4 is largely unchanged from last quarter. Looking ahead to next year, there will be dynamics that we will have to navigate, but we are confident in our ability to secure supply and adjust pricing as needed."

Kennedy also touched on the fiscal 2027 outlook. He said: "We have strong conviction in our AI business, supported by what we see in our backlog, the pipeline, and ongoing customer discussions. We’ve proven we can execute and deliver for our customers in this space."
 

Larry Dignan

Editor in Chief of Constellation Insights Constellation Research

About Larry Dignan: Dignan was most recently Celonis Media’s Editor-in-Chief, where he sat at the intersection of media and marketing. He is the former Editor-in-Chief of ZDNet and has covered the technology industry and transformation trends for more than two decades, publishing articles in CNET, Knowledge@Wharton, Wall Street Week, Interactive Week, The New York Times, and Financial Planning. He is also an Adjunct Professor at Temple University and a member of the Advisory Board for The Fox Business School's Institute of Business and Information Technology.&nbsp; <br>Constellation Insights does the following: Cover the buy side and sell side of enterprise tech with news, analysis, profiles, interviews, and event coverage of vendors, as well as Constellation Research's community and… Read more

An IBM executive said the company's steady and practical approach to quantum computing will win out over the bluster that's emerging from multiple vendors.

Speaking at an investor conference, Ric Lewis, senior vice president at infrastructure for IBM, was asked about Big Blue's approach to quantum computing. Lewis said quantum computing isn't about pumping out press releases as much as it is practical use cases and believable roadmaps.

"We are taking a very practical, rational approach to it," said Lewis. "We're not expecting some scientific breakthrough at this point. It's a matter of engineering and execution to get where we need to go."

• IBM, Cisco aim to scale, network quantum systems
• IBM launches IBM Quantum Nighthawk processor
• Quantum computing pure plays duel with giants, rivals

Lewis added:

"When I watch other quantum people and what they're saying I watch for a few things. One, do they have a believable roadmap. Not a roadmap, not just an aspiration but have you shown through your progress over the last five years that's you're on a certain trajectory. Do you have a believable roadmap for the next several steps?"

IBM's roadmap revolves around delivering quantum advantage in the next year and error correction and other capabilities later in the decade. "In '28, '29 we believe we'll be transacting on a system kind of level," said Lewis. "We're already transacting and we have clients that are buying cycles of quantum."

Lewis said IBM's roadmap is predictable and features an ecosystem of partners and a software stack. IBM features Qiskit, quantum computing software that has solid adoption.

Holger Mueller, an analyst at Constellation Research, said:

"Technical breakthroughs in commercial production do not happen overnight, but are the result of a string of successful completions of architectural advances that get delivered on time and functionally working. That is what IBM has done over the last 2-3 years. That is the progress and fidelity enterprises want to see when adopting a new technology platform is no exception."

Overall, Lewis said quantum computing is fragile and systems will need more resilience. He argued that combining classic and quantum techniques for error correction in quantum computing will be a practical approach to solving big problems.

• IBM Boosts the Software Side of Quantum
• IBM Embarks on a Fault-Tolerant Quantum Voyage
• Constellation ShortListâ„¢ Quantum Computing Platforms
• Constellation ShortListâ„¢ Quantum Full Stack Players

"I also look for a philosophy that says quantum is not a replacement to classical," said Lewis. "When you combine them together, you end up with something very strong. And since we play strong in classical, we play strong in AI, and I think we're the leader in quantum, we're really well positioned for as the industry gets to this kind of 2030 time frame and all that TAM. So we're pretty bullish and excited about it, though cautious and practical. Just keep executing the road map, make our steps, and we're going to be in a really good spot."

My take

Lewis has a point regarding the bluster meter in quantum computing. I'll tend to listen more to executives that refrain from the trash talk.

The reality is deployments at scale and returns on investment are probably five years away. Compute, networking and hybrid HPC-quantum systems are in the nascent stages. If you listen to quantum executives and the leading players, most companies are talking the same timelines.

Bluster has led to bigger valuations for some of the pure plays and enabled them to build strong balance sheets despite paltry sales. I'm willing to bet we're entering a new quantum computing phase where being more understated plays better. In the end, you need to deliver the qubits, error correction, software stack and scalability over the press release count.

Larry Dignan

Editor in Chief of Constellation Insights Constellation Research

About Larry Dignan: Dignan was most recently Celonis Media’s Editor-in-Chief, where he sat at the intersection of media and marketing. He is the former Editor-in-Chief of ZDNet and has covered the technology industry and transformation trends for more than two decades, publishing articles in CNET, Knowledge@Wharton, Wall Street Week, Interactive Week, The New York Times, and Financial Planning. He is also an Adjunct Professor at Temple University and a member of the Advisory Board for The Fox Business School's Institute of Business and Information Technology.&nbsp; <br>Constellation Insights does the following: Cover the buy side and sell side of enterprise tech with news, analysis, profiles, interviews, and event coverage of vendors, as well as Constellation Research's community and… Read more

Alibaba's cloud revenue in the second quarter surged 34% driven by AI workloads. Alibaba's cloud revenue is on an annual revenue run rate is more than $22 billion.

The Chinese retail and cloud giant said second quarter revenue for its Cloud Intelligence Group was $5.59 billion, up 34% from a year ago. Earnings before taxes, interest and amortization were $506 million in the second quarter.

Alibaba said it is investing heavily on building its vertical AI stack. The company is also betting heavily on open source large language models. As of Oct. 31, Alibaba's open source Qwen models have led to more than 180,000 derivative models on Hugging Face.

The company added that it is seeing "accelerating adoption of our AI products across a broad range of enterprise customers, with a growing focus on value-added applications including coding assistants."

Alibaba said it will continue to invest in AI products and its AI infrastructure.

In September, Alibaba Cloud outlined upgrades to its stack including new servers, networking, distributed storge and computing clusters. That infrastructure complements what Alibaba calls Platform for AI and model training and inference services.

Alibaba Cloud's stack includes:

• And upgraded set of databases, containers, storage and compute services optimized for data and AI workloads.
• Qwen3 family of models including Qwen3-Max, which has instruct and thinking versions, and Qwen3-Omni, a multilingual and multimodal model.
• The upcoming Wan2.5 video generation model.
• An upgraded agent development and application platform led by Model Studio.
• Model-Studio-ADK (Agent Development Kit) designed enterprise use cases.
• An upgraded AgentBay, a multimodal cloud operating environment and expert agent platform.
• AgentOne, an enterprise AI application platform.

Larry Dignan

Editor in Chief of Constellation Insights Constellation Research

About Larry Dignan: Dignan was most recently Celonis Media’s Editor-in-Chief, where he sat at the intersection of media and marketing. He is the former Editor-in-Chief of ZDNet and has covered the technology industry and transformation trends for more than two decades, publishing articles in CNET, Knowledge@Wharton, Wall Street Week, Interactive Week, The New York Times, and Financial Planning. He is also an Adjunct Professor at Temple University and a member of the Advisory Board for The Fox Business School's Institute of Business and Information Technology.&nbsp; <br>Constellation Insights does the following: Cover the buy side and sell side of enterprise tech with news, analysis, profiles, interviews, and event coverage of vendors, as well as Constellation Research's community and… Read more

Zoom reported better-than-expected third quarter results as the company gained wallet share and grew the number of customers contributing more than $100,000 in trailing 12 month revenue.

The company reported third quarter net income of $612.9 million, or $2.01 a share, on revenue of $1.23 billion, up 4.4% from a year ago. Non-GAAP earnings per share in the quarter was $1.52.

Wall Street was expecting Zoom to report earnings of $1.44 a share on revenue of $1.21 billion.

Zoom continued to show more growth in its enterprise segment. Enterprise revenue was $741.4 million, up 6.1% from a year ago, and online revenue, which usually refers to small businesses and consumers, had revenue of $488.4 million, up 2%.

• Zoom launches AI Companion 3.0, bevy of CX tools

By the numbers:

• Zoom reported that it had 4,363 customers contributing more than $100,000 in trailing 12 months revenue.
• Online monthly churn in the third quarter was 2.7%, or flat from a year ago.
• Workvivo had 1,225 customers, up 70% from a year ago.
• Zoom Phone passed 10 million paid seats in the third quarter.
• Zoom ended the quarter with $7.9 billion in cash and equivalents.

In prepared remarks, CEO Eric Yuan said the company is seeing strong demand for its AI Companion. He said team chat monthly active users were up 20% from a year ago. Zoom’s AI Companion integrates with Google Workspace, Microsoft 365 and Teams, Slack, Salesforce and ServiceNow.

"AI isn’t just bolstering our core, it’s opening new revenue streams and deeper customer value through customization and automation. Two quarters in, Custom AI Companion is scaling with several Fortune 200 wins and broad interest," said Yuan.

He noted that customer experience is gaining due to AI. "Within Customer Experience, AI has become a clear differentiator, creating additional monetization opportunities. Nine of our top ten CX deals involved paid AI, such as Zoom Virtual Agent or AI Expert Assist, as enterprises use Zoom to deliver faster, more personalized service," Yuan said.

As for the outlook, Zoom projected revenue between $1.23 billion and $1235 billion with non-GAAP earnings between $1.48 a share and $1.49 a share.

For fiscal 2026, Zoom said revenue will be between $4.852 billion and $4.857 billion with non-GAAP earnings of $5.95 a share and $5.97 a share.

Larry Dignan

Editor in Chief of Constellation Insights Constellation Research

About Larry Dignan: Dignan was most recently Celonis Media’s Editor-in-Chief, where he sat at the intersection of media and marketing. He is the former Editor-in-Chief of ZDNet and has covered the technology industry and transformation trends for more than two decades, publishing articles in CNET, Knowledge@Wharton, Wall Street Week, Interactive Week, The New York Times, and Financial Planning. He is also an Adjunct Professor at Temple University and a member of the Advisory Board for The Fox Business School's Institute of Business and Information Technology.&nbsp; <br>Constellation Insights does the following: Cover the buy side and sell side of enterprise tech with news, analysis, profiles, interviews, and event coverage of vendors, as well as Constellation Research's community and… Read more

Amazon said it will invest $50 billion to build and deploy AI and high performance computing AWS infrastructure built for the US government.

Construction on the infrastructure will break ground in 2026. The new investment will add almost 1.3 gigawatt of compute capacity for AWS Top Secret, AWS Secret and AWS GovCloud Regions for multiple classification levels.

Hyperscale cloud providers have been investment multiple billions of dollars to build out capacity for AI. These cloud providers also see a big total addressable market for public sector and government customers.

• OpenAI inks $38 billion deal with AWS, starts renting GPUs immediately
• AWS re:Invent 2025 to feature Trainium3, scaling secure AI agents
• AWS fires up Project Rainier, Trainium2 cluster for Anthropic

According to AWS, which kicks off its re:Invent 2025 conference next week, the US government focused infrastructure will be powered by its AWS Trainium AI chips as well as Nvidia GPUs. Federal agencies can use the infrastructure to access Amazon SageMaker AI, Amazon Bedrock and multiple models including Anthropic's Claude and Amazon Nova.

Use cases for the new infrastructure will range from national security to scientific research and innovation. AWS has been building US government focused clouds and infrastructure since 2011.

AWS CEO Matt Garman said the company's first-ever supercomputing infrastructure for government customers will give agencies "expanded access to advanced AI capabilities."

Constellation Research analyst Holger Mueller said:

"AWS has the largest government share of the cloud providers, and as such has practically the most to lose. Implicitly AWS confirms that HPC is a potential disruptor for commercial relations of cloud vendors - and therefore has to invest in it. The key question is going to be if AWS can pony up the capex for other key hardware innovations, AI super computers first, quantum platforms next."

Larry Dignan

Editor in Chief of Constellation Insights Constellation Research

About Larry Dignan: Dignan was most recently Celonis Media’s Editor-in-Chief, where he sat at the intersection of media and marketing. He is the former Editor-in-Chief of ZDNet and has covered the technology industry and transformation trends for more than two decades, publishing articles in CNET, Knowledge@Wharton, Wall Street Week, Interactive Week, The New York Times, and Financial Planning. He is also an Adjunct Professor at Temple University and a member of the Advisory Board for The Fox Business School's Institute of Business and Information Technology.&nbsp; <br>Constellation Insights does the following: Cover the buy side and sell side of enterprise tech with news, analysis, profiles, interviews, and event coverage of vendors, as well as Constellation Research's community and… Read more

NATO will deploy Google Distributed Cloud, an air-gapped version of Google Cloud on-premise.

Under a multi-million dollar contract with NATO's Communication and Information Agency (NCIA), Google will deliver sovereign cloud services to NATO for edge computing and AI use cases.

GDC is a sovereign cloud in a box that’s physically disconnected from the internet but includes everything to deploy virtual machines, run workloads and use services such as Vertex AI.

Google Distributed Cloud (GDC) was a big theme at Google Public Sector's annual conference in Washington DC in October. GDC was being used in public sector deployments, defense use cases and sovereign data and cloud implementations. Typically, GDC was being used in edge locations with limited to no connectivity and hardened environments.

• Google Public Sector Summit 2025: Takeaways on data, AI, ROI from 7 technology leaders
• Google Public Sector: AI agents and the future of government
• Google Public Sector, Lockheed Martin pair up for on-premises AI

In a statement, GDC will support NATO Communication and Information Agency's Joint Analysis, Training and Education Centre, which is using the infrastructure to modernize and manage classified workloads.

With GDC, NATO will maintain data residency and operational controls as well as autonomy. NATO has been building an interoperable communications and information systems architecture that can bring together 12 allied nations and 36 NATO entities quickly and at scale. The goal is to enable forces to operate together on the fly.

Constellation Research's Holger Mueller said:

"Google Cloud has for a long time invested in its next-generation computing -platform - originally Anthos - now Google Distributed Cloud (GDC). The key selection criteria for a next-gen computing platform is workload portability between public clouds and on premises. The required characteristic is identicality. The tech stack in the supported deployment is identical, portable and has investment protection of code assets. With Gemini and Vertex AI running on GDC, Google sets itself apart from the other next-gen computing vendors, and is therefore very well secured for air gapped solutions required for military workloads."