Logo
Become a Client

The story being written by the AI market is extraordinary by almost any measure. OpenAI’s revenue grew from $2 billion in 2023 to $20 billion in 2025. Anthropic’s annualized revenue run rate surged from $87 million in January 2024 to $30 billion by April 2026, a trajectory that Salesforce took 20 years to achieve. NVIDIA’s revenue grew eightfold from $27 billion to over $216 billion between 2023 and 2026, achieving that growth in just one-third of the time it took Apple to do so during its heyday between 2007 and 2015. By nearly every reported metric, the AI market is delivering growth at a scale and speed that defies historical comparison.

Yet a narrative has emerged about how much of that growth reflects genuine market demand and how much reflects the circular financing structures that have become a defining feature of AI infrastructure investment. Circular financing in AI describes investment structures where the same capital flows simultaneously as vendor payment and equity stake. A company funds its own customer’s revenue while also supplying that customer’s core infrastructure. The result is reported revenue growth that is real but not cleanly separable from investment activity. The answer matters enormously, not just for investors in AI infrastructure companies, but for every enterprise software vendor and technology buyer trying to make sense of what’s actually happening in the market.

Without getting into any financial analysis of the diagram below, I include it simply to demonstrate the extreme complexity and circular nature of these financial relationships.

Figure 1:

The circular financing problem

Circular financing in AI works through a self-reinforcing loop. NVIDIA committed $30B as part of OpenAI’s $110B financing announced in February 2026, while simultaneously serving as OpenAI’s primary GPU supplier, making it both a major investor in and a vendor to the same company. NVIDIA simultaneously holds equity in CoreWeave, which supplies infrastructure to Oracle, which signed a $300 billion Stargate commitment with OpenAI. Microsoft has invested more than $13 billion in OpenAI while serving as its primary cloud provider, meaning a substantial portion of OpenAI’s rapidly escalating compute spend flows back into Azure revenue. Microsoft has disclosed more than $600 billion in AI-driven remaining performance obligations, of which management confirmed approximately 45% is attributable to OpenAI-related activity. These are just some of the interdependencies that exist, and while they do not make the revenue fabricated, they do make it very difficult to read reported growth figures as clear evidence of external market demand expanding at the rates that match recent headlines.

Stripping out circular flows to estimate genuine arm’s-length revenue is nearly impossible from the outside because none of these companies has any incentive to disaggregate them. What is clear is that OpenAI’s compute costs, projected to reach tens of billions annually, still dramatically exceed its current revenue. While their internal projections point to massive revenue growth over the longer term, the company is expected to remain unprofitable through at least 2029, with positive cash flow not expected until 2030. The infrastructure layer is being built on a combination of genuine demand and financial engineering, and the two are not currently separable from reported figures alone.

Why the application layer is the real signal

This is why revenue growth on the enterprise application layer has become the most important signal in the entire AI market. Enterprise application software does not carry a circularity problem. When an ERP vendor reports AI-driven ARR expansion, when an HCM platform demonstrates higher attach rates on AI-enabled capabilities, when a finance or procurement solution delivers measurable process efficiency gains that a CFO chooses to fund again at renewal, those signals reflect real buyers making real budget decisions based on perceived business value. There is no investment web distorting the demand signal. The revenue either reflects a genuine willingness to pay for demonstrated outcomes, or it does not.

What’s interesting is that we’ve seen this same exact pattern play out multiple times before during historical tech market revolutions. In each of the last several significant platform transitions, the application layer lagged the infrastructure layer in value creation, but eventually surpassed it as the platform matured.

The historical pattern

During the internet buildout, infrastructure companies captured the overwhelming share of market value. Cisco dominated with networking equipment and briefly became the most valuable company on earth, reaching a market cap of nearly $560 billion in March 2000. Companies like Sun Microsystems and Dell supplied servers, and even fiber optic cable makers like Corning and JDS Uniphase had their moment. Meanwhile, the application layer was largely characterized by money-losing dot-com ventures that collapsed when the bubble burst. But over the long term, application layer players like Amazon, Salesforce, and Google eventually dwarfed those companies.

The cloud era was no different. AWS launched in 2006 and, for years, held a dominant share of the cloud market, while enterprise SaaS applications were still proving their business model viability. IDC’s Worldwide Semiannual Public Cloud Services Tracker shows that by 2020, the SaaS application layer had grown to $148 billion in revenue, nearly half of the total $312 billion public cloud market, while IaaS, despite its faster growth rate, remained a significantly smaller share of the overall pie. Once the cloud infrastructure platform matured, value creation in the application layer far exceeded it.

Looking at the current AI boom, it’s following the same early-stage script. NVIDIA first became the world’s most valuable company in mid-2024, and we are in the middle of an AI infrastructure super cycle with trillions of dollars flowing into compute, datacenter construction, and power generation. Meanwhile, enterprise AI application revenues remain nascent relative to the trillions being invested in the infrastructure layer beneath them.

Agentic AI is triggering a fundamental reconsideration of how companies invest in packaged software, as AI agents effectively become the new enterprise apps. IDC’s spending forecasts project AI investment growing 31.9% annually between 2025 and 2029, reaching $1.3 trillion. These are real demand signals, and they are large enough to sustain significant growth at the application layer if enterprise software vendors can connect their AI capabilities to the business outcomes buyers are seeking.

The monetization gap that still needs closing

The monetization gap, however, remains significant and is not yet closing at the pace the market requires. IDC’s Future Enterprise Resiliency and Spending Survey (FERS) finds that few organizations report measurable financial results from their AI projects, despite widespread improvements in individual productivity. While most organizations aspire to grow revenue through AI initiatives, the majority have yet to achieve that goal. This distinction between individual productivity gains and organizational-level financial outcomes is one that enterprise software vendors have been slow to confront directly. Productivity at the individual level is real, documented, and genuinely valuable. Translating those gains into measurable business results that justify enterprise software pricing premiums is a fundamentally different challenge, and one that buyers are increasingly demanding vendors address more explicitly during renewal and contract-expansion discussions.

What vendors must do now

Closing that gap is the defining commercial challenge for enterprise software vendors over the next 12 months. The most urgent shift is from feature availability to outcome accountability: specific, auditable improvements in process efficiency or cost reduction that survive CFO scrutiny at renewal. Transparent commercial models follow from that. If buyers have to build the business case themselves, or fund significant professional services to reach value, adoption curves will flatten and renewal risk will rise. The competitive differentiator in this environment is not model sophistication. It is workflow transformation. Enterprises evaluate software on whether it changes the workflows that actually drive their business.

The infrastructure layer of the AI market may ultimately prove resilient. The circular financing dynamics, while real, are not unlike the capital formation structures that funded prior technology infrastructure buildouts, and genuine demand for compute at scale is not in dispute. The application layer is not a downstream beneficiary of the AI infrastructure boom. It is the proof point on which the entire narrative depends.

Eric Newmark

Eric Newmark - Group Vice President & General Manager of IDC's SaaS, Enterprise Software, CX and Workplace Solutions Division

Analyst Profile
Eric Newmark is Group Vice President & General Manager of IDC’s SaaS, Enterprise Software, CX, and Workplace Solutions Division, which includes several teams of analysts covering SaaS, 18 enterprise application markets, software monetization, business platforms, marketplaces, and services firms focused…

The AI assistant era is wrapping up.

Chatbots, copilots, summarizers—useful, yes, but that’s the warm-up act. What comes next is agentic AI: systems that don’t just respond to prompts but plan, reason, and execute work across your systems, data, and workflows, with or without human augmentation. Executives are putting agentic AI at the center of their business transformations, reimagining how people work with AI-powered workflows.

By 2029, IDC projects 1.15 billion active agents, 40 times the 2025 baseline, executing 217 billion actions per day.

The stakes are not theoretical

Business models, skills, economics, governance, compliance, and security requirements are changing rapidly—creating massive, rapidly evolving risks. The numbers bear that out:

CEOs, board members, and shareholders are demanding AI results now. Waiting will stifle market share, growth opportunities, and innovation. Early winners will widen their advantage by applying proven AI learnings across new use cases. In the AI economy, continuous learning creates speed and compounding results.

This isn’t an upgrade. It’s a new operating model

Where a copilot drafts an email, an agentic agent manages the entire workflow behind it: pulling data, making decisions, escalating exceptions, closing the loop. Autonomously. Persistently. Across systems. With or without human intervention. This requires executives to redesign processes with AI at the center, not bolted on, but as the heartbeat of every workflow.

IDC finds that by 2027, enterprises will need architectures that support synchronous movement of data, apps, workflows, and agents. If your current stack wasn’t designed for that, the clock is already running.

“Agentic systems introduce a fundamentally different risk profile than the AI tools most enterprises deployed in 2023 and 2024.”

The implications extend beyond architecture. Copilots operated under human supervision—every output reviewed, every action confirmed. Agents operate differently: they initiate, chain decisions, and act across systems in real time. A misconfigured agent doesn’t just produce a bad answer; it executes a bad decision, potentially at scale, before anyone notices. That shift from human-reviewed output to autonomous action is where governance frameworks built for the assistant era break down entirely.  Governance, security, and orchestration are critical investment requirements, not an afterthought for effective and efficient AI outcomes. Governance and security must be built in from day one, the business and career risks are too high to treat them as afterthoughts. Every digital worker will be governed like a human employee.

Executives who recognize this distinction early hold a structural advantage. They’re not just deploying faster—they’re building the operational muscle, audit infrastructure, and trust frameworks required to scale safely. Those who treat agentic AI as a continuation of their copilot strategy will face compounding remediation costs as agent portfolios grow, and governance gaps widen.

The competitive divide forming now isn’t between companies that have AI and those that don’t. It’s between organizations that have learned to deploy agents with discipline and those still improvising. Speed without structure is exposure. Structure without speed is irrelevance. The executives navigating both are the ones establishing the terms. 

Seven domains. One framework

IDC’s deployment framework surfaces seven domains where early decisions have outsized downstream impact. Get them right and you build momentum:

  • Business outcomes
  • Workflow design
  • Data readiness
  • Runtime architecture
  • Governance
  • Operating model
  • Workforce

Each domain highlights key considerations for executives to gauge their deployment and governance readiness. After the self-assessment, the research provides deep dives into critical areas of transformation—software development, partner and AI infrastructure selection, and IT operations. These recommendations serve as a discussion guide for your agentic AI journey.

Key takeaways for executives

  • Treat agentic AI as a new operating model, not an upgrade to your copilot strategy.
  • Governance and security are now a competitive differentiator—not an IT compliance exercise.
  • The 7-domain framework is your cheat sheet: use it to reduce deployment risk and extract compounding value.
  • Early movers will set the terms. Waiting isn’t neutral—it’s a strategic concession.

Navigate your next move with confidence: Download the full IDC research, Leading Through the Agentic Deployment Era, for the complete self-assessment framework, domain deep dives, and actionable deployment recommendations.

Stephen Elliot

Stephen Elliot - Group Vice President, I&O, Cloud Operations, and DevOps

Analyst Profile
Stephen Elliot manages multiple programs spanning IT Operations, Enterprise Management, ITSM, Agile and DevOps, Application performance, Virtualization, Multi-Cloud Management and Automation, Log Analytics, Container Management, DaaS, and Software Defined Compute.   Mr. Elliot advises Senior IT, Business, and Investment Executives globally in the creation of…

Strip away the keynote theatrics and the RTX Spark launch at Computex comes down to a single wager: a meaningful share of future AI work belongs on the PC rather than in a data center. NVIDIA and Microsoft built a Windows platform around a strong GPU and a deep pool of unified memory because they want models and agents running where the user sits. For the next year or two, open-weight models are what make that wager credible. Over a longer horizon, the bet only pays off if the companies behind the best frontier models are willing to change how they do business.

To date, the Windows AI message has leaned on the neural processing unit (NPU) and the idea of small, persistent AI workloads carried out locally. But the early returns were thin. Copilot+ shipped with fewer headline features than promised, and developers never showed up to write for the NPU in any meaningful numbers. It says something, then, that this launch put a GPU and an agent runtime at the center, and went silent on the NPU and on Copilot+ branding. Both will presumably be present in these as-yet-released machines. The choice to keep them offstage is the part worth noting.

Governance first, compute second

Most of the engineering that should interest corporate buyers has little to do with raw compute. NVIDIA and Microsoft added containment and identity controls to Windows and wrapped them in an agent layer called NVIDIA OpenShell. An administrator can spell out exactly what an agent may reach, push sensitive requests to a model running on the machine, and scrub identifying details before any prompt leaves for the cloud. IT teams have made clear they will not let autonomous software roam across managed devices without that kind of governance. Whether it works as advertised is unproven, but the design attempts to answer ITDMs’ legitimate concerns.

Local inference = lower cost

The case for keeping inference on the device is mostly about money. Privacy, security, and lower latency matter too — but the key is that a generous memory pool attached to a robust GPU lets a capable model sit resident and respond locally. The work does not run up charges in a cloud instance every time an agent acts. Once agents are working all day underneath ordinary tasks, paying cloud rates for every token they burn stops being sustainable well before the technology hits any wall. Shifting a portion of that load onto hardware the buyer already owns is one of the few ways the forecasted agentic AI ramp proceeds unimpeded.

Today, open-weight models may be capable enough to run useful agents on a machine like the RTX Spark. That’s the near-term proof point. The best models, though, still sit behind cloud subscriptions from Anthropic, Google, and OpenAI. Companies trust those models, and the AI firms behind them have built large businesses on running inference in their own data centers. Moving part of that work onto a customer’s GPU asks them to do more than ship a smaller model — it asks them to redraw the business: pricing and packaging a hybrid where some inference runs locally and some runs in the cloud. Local tokens for the simple jobs, cloud tokens for the heavy lifting. None of the foundation model companies have offered a subscription like that yet. Until one does, on-device intelligence will run on open models that will likely always trail the frontier.

Price, timeline, and the real test

The practical limits are easy to see. No RTX Spark notebook prices have been set, and launch configurations will almost certainly sit at the top of the premium market. Average selling prices for notebooks have drifted upward for years, and that was before memory costs exploded. Today’s memory pricing will make a fully loaded 128GB RTX Spark system quite steep. The buyers willing to pay for the maximum build will mostly be developers and creative professionals who already understand what running models locally gets them. Everyone else is further out.

None of this makes RTX Spark the mainstream arrival of the agentic PC. The silicon looks capable, and adding a fourth serious contender alongside Intel, AMD, and Qualcomm should sharpen all of them. The shift to more AI compute on device is the more straightforward part of this equation. The questions that determine whether this matters are softer and slower: whether Microsoft’s agent platform earns the trust of cautious IT departments, and whether the frontier labs conclude that some of their inference belongs on the customer’s desk. Settle those two and RTX Spark will look ahead of its time.

Read the full analysis
Tom Mainelli

Tom Mainelli - Group Vice President, Device & Consumer Research

Analyst Profile
Tom Mainelli heads the Device & Consumer Research Group, overseeing a wide array of hardware and technology categories catering to both home and enterprise markets. His team's research spans PCs, tablets, smartphones, wearables, smart home devices, thin clients, displays, and…

Most EMEA organisations have the intent to scale AI. What they are missing is a way to execute. On May 28, 2026, IDC’s EMEA Digital Leaders Hub brought together Martina Longo, Daniel Saroff, and Giulia Carosella for a live session drawing on 12 months of conversations within the Hub and fresh IDC research. Below is a brief overview. The full recording is available on demand. 

AI maturity in EMEA in 2026: why execution, not intent, is the real gap 

IDC’s latest MaturityScape Benchmark (EMEA, N=583) tells a clear story: 63% of EMEA organisations are still in the two lowest AI maturity stages. Fourteen percent are ad hoc: scattered initiatives, no coherent strategy. Forty-nine percent are opportunistic, running pilots but without the repeatability needed to scale. Just 2% are effectively scaling AI and Agentic AI initiatives across their organizations, including unlocking AI-driven revenue growth. 

The journey maps from the (Gen)AI Scramble (fragmented, investment-heavy experimentation) through the AI Pivot (structured scaling) to the Agentic Organisation (AI embedded across operations). Most EMEA organisations are stuck in the transition between the first and second stage. The blocker is almost never ambition. It is the ability to execute. 

Why AI adoption in EMEA is stalling: five challenges organisations need to address 

Notably, 49% of EMEA organisations have already shifted their focus from launching new AI pilots to improving existing initiatives The experimentation phase is peaking, and EMEA organizations are no longer seeking new tools but instead focusing on making current AI work effectively first. But five structural challenges continue to slow progress: 

  • Competition for resources among digital initiatives 
  • Regulatory uncertainty slowing deployment decisions 
  • Resistance to process change within the business 
  • Difficulty quantifying and demonstrating AI ROI to the board 
  • Lack of executive sponsorship or organisation-wide buy-in 

These are not isolated problems. They compound each other. Without a shared language for value, the resource conversation is difficult to win. The webinar addressed all five, and the one that generated the most discussion was ROI. 

Measuring AI ROI: why cost savings are not enough 

The most common reason AI initiatives stall is not technical. It is that no one agreed upfront on what success looks like. IDC’s AI Business Value Benefit framework maps nine dimensions where AI creates measurable impact, spanning Revenue Generation and Customer Experience through to Sustainability, Time to Market, and Business Resilience. Most organisations are measuring only one or two of these dimensions and are therefore systematically underselling the value they already have. 

“Know what you want to achieve and how you will measure it” 
  — Alex Catmur, Commercial Director Digital, AtkinsRéalis 

Three practices that came up consistently in the session: 

  • anchor every initiative to a specific business outcome before selecting any tool: Start with value drivers, not technology 
  • if no executive has a stake in the metric, the initiative will eventually stall: Align to KPIs executives already own 
  • productivity gains are visible; resilience and trust are harder to quantify but equally real, and the framework accounts for both: Separate direct and indirect value 

The session also walked through a detailed case study of a global professional services firm that went from no shared ROI lens to confident scale decisions. If that is where your organisation is at present, it is worth watching the recording to hear how they structured the turnaround. 

How the CIO role is evolving in the age of AI 

IDC’s WW C-Suite Tech Survey (EMEA, N=300) makes the expectation clear: 42% of the broader C-suite now expects the CIO to lead digital and AI transformation with a major focus on creating new revenue streams. That expectation is growing faster than the formal authority that would make it achievable. 

The digital leader of the future, as IDC frames it, is an architect of three things: Workforce (orchestrating AI-fuelled change management), Resilience (modernising IT for strategic alignment), and Value (demonstrating what digital technologies actually deliver for the business). 

Three steps for digital leaders looking to prepare for this shift: 

  • not called in to implement choices that have already been taken: Get in the room before decisions are made 
  • accountability for results, not just go-live dates: Own the business outcome, not just the delivery 
  • design AI deployments to grow and withstand failure, not just to ship: Architect for scaling and recovery from the start 

The session went into considerably more detail on each of these steps, including the structural and political dynamics that make them harder in practice than they appear on paper. 

A practical AI transformation playbook: four steps that matter

The session closed with a synthesis that cuts through the complexity. Leading AI-fuelled business transformation comes down to four sequential actions: 

  • modernise architecture and data before scaling AI; pilots built on brittle infrastructure do not survive production: Fix the foundation 
  • anchor every initiative to a KPI an executive already owns; no metric, no mandate: Define the value 
  • redesign workflows and roles alongside the technology; AI layered onto old processes delivers expensive old results: Change the model 
  • run experiments to disprove bad assumptions quickly; promote only what survives to a funded pilot with a scale plan: Scale what works 

Straightforward to state, harder to execute. The webinar covers what this looks like in practice, including the Q&A that followed. 

► Watch the full recording on demand 
 

The complete recording covers the full AI Business Value Benefit framework across all nine dimensions, the case study of a professional services firm moving from pilot to scale, the detailed best practice sessions on ROI measurement and the evolving CIO mandate, and the Q&A with the IDC analysts. If the topics covered are relevant to your organisation’s AI journey, IDC’s EMEA Digital Leaders Hub offers advisory support, benchmarks, and peer roundtables for CIOs and digital leaders navigating this transition. Reach out via the contact form to continue the conversation. 

Martina Longo

Martina Longo - Research Manager, CIO and CTO Buyer Insights

Analyst Profile
Martina Longo is a Research Manager for the CIO and CTO Buyer Insights program. Her research focuses on the emerging priorities, programs and decision making processes linked to the modern CIO/CTO agenda. This includes advancing AI from experimentation to operational…
Daniel Saroff

Daniel Saroff - Group Vice President Research and Consulting

Analyst Profile
  Daniel Saroff is Group Vice President of Research and Consulting at IDC, where he leads the research agenda focused on end-user technology leaders, including CIOs and their direct leadership teams. He oversees a team of analysts and advisory professionals…
Giulia Carosella

Giulia Carosella - Senior Research Manager

Analyst Profile
Giulia Carosella is a Senior Research Manager in IDC's AI-Fueled Business Strategies team, leading the Worldwide research program. In this role, she researches current and emerging global trends in AI?driven business transformation, examining how organizations can reinvent themselves by leveraging…

What Happened in India’s PC Market in Q1 2026? 

India’s traditional PC market (encompassing desktops, notebooks, and workstations) delivered a standout performance in Q1 2026, with total shipments reaching 4.4 million units, representing 31.1% year-over-year growth, according to IDC’s Worldwide Quarterly Personal Computing Device Tracker. Notably, this marked the third consecutive quarter in which shipments exceeded the 4-million-unit threshold, a milestone that underscores the market’s sustained momentum. The growth came even as PC prices continued to rise due to component shortages. Nonetheless, this reflects strong demand across both consumer and commercial segments, fuelled by digital transformation, enterprise investment, and large-scale education procurement. 

Why It Matters 

The Q1 2026 results signal the resilience of India’s PC market in the face of structural pricing pressures. Here’s what stakeholders should take away: 

  • Vendors should limit dependency on education-led demand and monitor the durability of government procurement pipelines. ELCOT was a major volume driver this quarter, but such big volume projects are rare.   
  • Rising prices are accelerating premiumisation, but they also risk squeezing out price-sensitive buyers in the consumer and SMB segments if unchecked. 

Market Dynamics: What Drove the Outcome? 

Three converging forces shaped the quarter’s strong performance: 

  • ELCOT education project led to notebook surge: The execution of the Tamil Nadu government’s ELCOT education procurement programme was the single largest growth catalyst. It drove the education segment to an exceptional 455.2% YoY growth and propelled the overall notebook segment to 3.3 million units, or a 55.4% YoY increase. 
  • Enterprise investment providing commercial resilience: Sustained domestic and global enterprise spending supported 24.2% YoY growth in the enterprise segment. Workstations, driven by high-performance computing needs in engineering and design, rose 31.8% YoY. 
  • AI adoption and premium segment expansion: Rising demand for AI-capable notebooks helped drive the premium notebook segment (above US$1,000) up 70.1% YoY. Consumer demand for premium devices outpaced commercial, growing 92.4% versus 56.1% YoY. 

Segment Snapshot: Winners and Laggards 

Notebooks: 3.3 million units | +55.4% YoY | Majority of total shipments 

Workstations: +31.8% YoY | Driven by engineering, design, and data-intensive sectors 

Desktops: −13.5% YoY | Elevated prices led to order deferments and cancellations 

Premium Notebooks (>US$1,000): +70.1% YoY | Consumer: +92.4% | Commercial: +56.1% 

AI Notebooks (total): +96.1% YoY  

India PC Market at a Glance: Q1 2026 

  • Total shipments: 4.4 million units (+31.1% YoY) 
  • Third consecutive quarter above 4 million units 
  • Commercial segment: 2.8 million units | Education: +455.2% YoY | Enterprise: +24.2% YoY 
  • Consumer segment: +11.7% YoY | eTailer: +23.1% YoY | Traditional retail: +8.7% YoY 
  • Top five vendors: HP Inc. (#1), Acer Group (#2), Lenovo (#3), Dell Technologies (#4), ASUS (#5) 

Analyst Insight 

“Despite persistent upward pressure on PC prices due to rising component costs, particularly DRAM and GPUs, consumer demand has remained resilient. With prices having increased over the past two quarters and expected to rise further in the coming months, proactive and transparent communication from vendors, partners, and channel stakeholders has encouraged early purchase decisions, thereby helping sustain overall market demand.” 

— Bharath Shenoy, Research Manager, Devices Research, IDC 

Vendor Landscape: Who Won the Quarter? 

HP Inc. retained the top spot with 26.6% market share, driven by ELCOT execution and solid enterprise demand (+27% YoY in commercial). Consumer growth was steady at 8.2% YoY, supported by strong large format retail (LFR) traction. 

Acer Group surged to second place with 21.3% market share, overtaking both Lenovo and Dell. ELCOT billing in the commercial segment, where it captured 25.5% share, was the key driver. A marginal 1.2% YoY dip in consumer shipments due to entry-level supply constraints was the only blemish. 

Lenovo held third with 18.3% market share. Commercial grew 32.3% YoY on enterprise and SMB strength, while consumer grew 20.8% YoY, bolstered by a gaming portfolio surge and e-commerce shipments up 58% YoY. 

Dell Technologies ranked fourth at 15.9% market share. Enterprise demand drove a 37.5% YoY segment increase, with overall commercial growing 38.7% YoY partly from ELCOT fulfilment. Consumer shipments grew 17% YoY despite pricing pressures. 

ASUS rounded out the top five at 6.9% market share. Gaming notebook demand lifted consumer growth 31.9% YoY, and an exceptional 241.1% YoY commercial surge reflected rising SMB momentum. 

IDC Outlook: What’s Next? 

The first half of 2026 is expected to sustain positive momentum, supported by favourable supply allocations, aggressive channel stocking, and front-loaded procurement ahead of anticipated price hikes. However, the second half presents a more cautious picture. 

  • What could sustain growth? Strong enterprise and SMB pipelines would drive 2Q 2026, continued AI notebook adoption, and successful rollout of products like Apple’s mass-market MacBook Neo (launched March 2026) might be some positive drivers for 2H 2026. 
  • What could slow it down? Inventory corrections in Tier 1 and Tier 2 channels, supply-side constraints, rising DRAM and GPU costs, and budget pressures in government and education segments. 
  • What should readers watch next quarter? Whether channel inventory levels normalise, how consumer demand responds to further price increases, and the pace of AI notebook adoption in SMBs. 

“Vendors have so far benefited from favourable supply allocations and are currently managing relatively high inventory levels across Tier 1 and Tier 2 channels. While enterprise and SMB segments continue to support volume-led procurement, shipments might decline in second half of the year. The government and education sectors could face challenges amid rising device prices and constrained budget allocations. The consumer segment may also witness a decline in the second half of the year unless pricing stabilizes. At the same time, Apple’s launch of the mass-market-focused MacBook Neo in March is expected to strengthen its position and potentially expand its presence across both consumer and SMB segments.” 

— Navkendar Singh, AVP, Devices Research, IDC 

Frequently Asked Questions 

Why did the PC market grow so strongly despite rising prices? 

The primary catalyst was the ELCOT education procurement project, which alone drove the education segment up 455.2% YoY and turbocharged notebook shipments. This institutional demand, combined with enterprise investments and early purchasing by both enterprises and consumers ahead of further price increases, more than offset the headwinds from elevated pricing. 

What risks could impact the PC market in 2026? 

The key risks include prolonged component cost inflation (especially DRAM and GPUs), potential inventory overhang in distribution channels, and weakening budget availability in government and education. If prices continue to climb, consumer spending moderation could also temper growth. 

About IDC 

International Data Corporation (IDC) is the premier global provider of trusted technology intelligence, advisory services, and events. With more than 1,000 analysts worldwide, IDC offers global, regional, and local expertise on technology, IT benchmarking and sourcing, and industry opportunities and trends in over 100 countries. IDC’s analysis and insights help IT professionals, business executives, and the investment community to make fact-based technology decisions and to achieve their key business objectives. To learn more about IDC, please visit www.idc.com. Follow IDC Asia/Pacific on X at @IDCAP and LinkedIn. Subscribe to the IDC Blog for industry news and insights. 

All product and company names may be trademarks or registered trademarks of their respective holders. 

For more information, contact Bharath Shenoy, Research Manager — Devices Research, or Navkendar Singh, AVP – Devices Research. 

Bharath Shenoy

Bharath Shenoy - Senior Market Analyst

Analyst Profile
Bharath Shenoy is a Senior Market Analyst at IDC India. Based in Bangalore, Bharath is responsible for tracking and analyzing market trends of PCs and Printers for Bangladesh. Prior to joining IDC, Bharath has worked with Dataxis as a Telecom…
Navkendar Singh

Navkendar Singh - Associate Vice President, Client Devices & IPDS, IDC India

Analyst Profile
Navkendar Singh is a Associate Vice President with IDC India, based in Gurgaon. His research domains encompass deep-dive research and insights in and around mobile devices, smart homes, PCs, tablets, wearables, and the printing market in India, Bangladesh, and Sri Lanka.…

The global PC market is heading into a turbulent second half of 2026, and there’s no quick fix on the horizon. IDC now forecasts global PC shipments will decline 11.3% for the full year, with conditions worsening progressively through Q4, when shipments are expected to fall 20% year-over-year.

The culprit is a persistent memory shortage with no meaningful relief expected before the end of 2027. The knock-on effects are significant: prices are rising and PC manufacturers are struggling to maintain full product portfolios.

Pull-forward demand masked the problem, temporarily

The first quarter of 2026 offered a deceptively encouraging signal, with shipments growing 3% versus the same period last year. But that strength was largely borrowed from the future. Buyers, both consumer and commercial, accelerated purchases ahead of anticipated price hikes and product availability constraints.

“The anticipation of rising prices and of limited availability of some system configurations due to memory and other component shortages has led some end-users to make purchases earlier than anticipated,” said Jean Philippe Bouchard, Vice President of Devices and Consumers at IDC. “We’re not seeing any relief to the memory shortage situation before the end of 2027, which means prices will continue to rise and PC manufacturers will struggle to maintain full product portfolios for the foreseeable future.”

Some of that momentum is carrying into Q2, but the remaining quarters are expected to deteriorate gradually and sharply.

The MacBook Neo factor

One notable wildcard is Apple’s MacBook Neo, which has driven stronger-than-expected notebook demand and prompted IDC to revise its notebook forecast upward. But its ripple effects cut both ways.

“The introduction of the MacBook Neo is putting real pressure on the entire PC ecosystem,” said Jitesh Ubrani, research manager for IDC’s Consumer Devices Trackers. “We expect vendors to respond with a combination of new silicon, a more efficient OS from Microsoft, and aggressive promotional pricing.”

“The competitive pressure from the Neo is providing a partial offset to broader price increases, keeping some low-cost notebook options alive. But the overall trajectory for average selling prices (ASPs) is firmly upward. IDC forecasts ASP growth of 17% in 2026, and even as memory capacity expands over the next two years, pricing is unlikely to return to 2025 levels,” Ubrani added.

What to watch

The PC market is navigating a rare combination of structural supply constraints, macroeconomic headwinds, and platform-level disruption. For buyers, the window for favorable pricing is narrowing. For vendors, the challenge is sustaining portfolio breadth while managing component scarcity and competitive pressure from Apple’s latest hardware.

IDC tracks ongoing developments through its Worldwide Quarterly Personal Computing Device Tracker.

Jitesh Ubrani

Jitesh Ubrani - Director, Consumer Devices Research

Analyst Profile
Jitesh Ubrani is a Director at IDC leading a team of analysts within the Worldwide Consumer Device Trackers group, covering wearables, augmented reality (AR), virtual reality (VR), tablets, phones, PCs, gaming, and smart home devices, with a focus on market…
Jean Phillippe Bouchard

Jean Phillippe Bouchard - Vice President, Data & Analytics

Analyst Profile
Jean Philippe (JP) Bouchard is Vice-President, Data & Analytics at IDC Canada. In this role, JP is responsible for leading the team of analysts delivering Continuous Intelligence Services, Trackers and custom research in the Future of Work and Mobility group,…

Fifty percent of organizations are already deploying AI agents in production across multiple business areas. Another 27% have them running in at least one. IDC’s latest Market Glance on agentic AI ecosystems maps the landscape that is making this shift possible. The picture it reveals is both more structured and more dynamic than many organizations currently appreciate.

A market built in layers

Think of the enterprise agentic AI market ecosystem as three interlocking layers. At the surface, there are agents themselves: specialized standalone agents built for specific domains, and in-application agents embedded directly inside the software your teams already use. These agents don’t need an external orchestration layer — they work within the host application’s environment, using its data and functionality to act autonomously on behalf of users.

Beneath that sits a rich set of creation and operation technologies. This is where organizations build, deploy, monitor, and manage agents throughout their full life cycle. It includes multiagent orchestration frameworks, agent discovery and inventory tools (so you know what you’ve deployed and where), and the data foundations (vector databases, knowledge graphs, real-time pipelines) that give agents the contextual grounding they need to work effectively.

Underpinning everything are the execution foundations: cloud and on-premises infrastructure, AI silicon from specialized semiconductor providers, and a growing set of security capabilities including AI security posture management, runtime policy enforcement, and identity and access management for agents. As organizations seek greater control over where and how AI runs, driven by privacy, regulatory and sovereignty concerns, architectural diversity in this layer is accelerating fast.

77% of organizations already have AI agents running in production. The question is no longer whether to deploy. It’s how to navigate what comes next. Read the report.

The convergence you need to understand

What’s making this market particularly dynamic right now is convergence. Enterprise software platform vendors and application providers are racing up the value stack, embedding agentic capabilities directly into their products. At the same time, professional services providers and systems integrators are moving in the opposite direction: building proprietary, asset-based delivery models anchored in the same underlying technologies.

The result is that buyers face a highly dynamic market space. IDC’s research shows organizations are evenly split between build and buy strategies, and most expect to source agents from multiple vendors simultaneously. That’s pragmatism. The agentic AI market is too young and too fast-moving for anyone to bet on a single supplier.

Your next move

Understanding where your organization sits in this ecosystem, and where your partners and competitors sit, is no longer optional. Agentic AI’s promise is the digitization of labor at scale. The organizations that navigate this landscape with confidence will be the ones that mapped it first.

IDC’s Market Glance gives business executives, product managers, and analyst relations professionals the structured view they need to assess opportunities, evaluate partnerships, and make the right decisions about where to place their bets. The ecosystem is moving quickly. Navigate your next move with the evidence to back it up.

Neil Ward-Dutton

Neil Ward-Dutton - Research VP, Agentic Automation & AI Technologies

Analyst Profile
Neil Ward-Dutton is Research VP, Agentic Automation & AI Technologies at IDC. In this role he guides IDC’s research agendas, and helps enterprise and technology vendor clients alike make sense of the opportunities and challenges across markets spanning core AI…

Five years ago, enterprise technology roadmaps stretched across half a decade. Then they shifted to three years. Today, many organizations are planning just 18 months ahead, and even that horizon is getting shorter. As AI accelerates innovation, the challenge is no longer access to technology. The challenge is applying it effectively in a world that is changing faster than most organizations can plan for.

Perficient is not waiting for the dust to settle.

The global consultancy, with more than 7,000 advisors, engineers, and designers, has deep expertise in heavily regulated sectors such as banking, financial services, insurance, healthcare and life sciences, as well as auto and industrials. Working as a strategic partner to Fortune 100 companies, Perficient helps organizations become more data-driven, make better decisions, and move faster without losing their footing.

Its approach is grounded in a particular kind of pragmatism.

“Our focus is really on helping our clients understand the world around them and be very pragmatic in their approach to leveraging technology,” said Eric Walk, Vice President, AI Data Platforms at Perficient, “cutting through the noise and the buzzword bingo.”

That means treating AI as a practical tool rather than a destination, applying it to specific problems with clear outcomes in mind, and working iteratively with clients so they can adapt as conditions change.

“That type of agile iteration where we’re moving in weeks, not years, is the key to effective adaptation to the way the world is changing around us,” Walk said.

Building intelligence that can keep up

Speed alone does not improve outcomes. Organizations also need confidence in the direction they are moving.

Perficient uses a balanced intelligence model that brings together proprietary knowledge, open-source intelligence, and premium analyst research to create a complete view of the market. AI helps teams analyze large volumes of information and surface insights more quickly, and the firm is building internal AI agents to help consultants work more efficiently and deliver higher-quality work for clients. But the quality of those outputs depends entirely on the quality of what goes in.

“You can open up the world and have AI crawl the internet and look at any source of information, but you’re going to get results that reflect the internet,” Walk said. “It’s critical for us to ensure we have trusted inputs to produce trusted outputs.”

The role of trusted insight

That focus on trusted inputs is where analyst research earns its place in the model.

IDC’s research helps Perficient teams validate assumptions, challenge internal perspectives, and understand how markets are evolving in real time. It fills gaps in areas where consultants may not have deep specialization and gives them the external grounding to give clients advice they can stand behind.

“The more curated, more insightful, more thoughtful content that IDC provides is critical in helping us give our clients more effective advice and understand the world that we and they are living in day to day,” Walk said.

As AI becomes more integrated into consulting workflows, how teams access that insight is evolving too. Walk was among the beta testers for IDC Quanta, IDC’s new AI platform built on its proprietary research. For a firm already invested in making consulting more natively AI-enabled, the ability to query trusted IDC intelligence through a conversational interface, with every response grounded in specific research, points directly at the problem Perficient is trying to solve.

Turning disruption into advantage

Perficient is applying these same principles to itself, actively building AI-driven capabilities to improve how its own teams operate across research, delivery, and execution. Whatever it recommends to clients, it is already doing internally.

“We want to be out front and thinking about new ways of doing business and putting them into practice to drive outcomes for our clients,” Walk said. “At the end of the day, what matters is increasing revenue, reducing cost, and driving efficiencies in the right way to make both us and our clients successful.”

In a market that is not slowing down, that combination of trusted intelligence, honest advice, and a genuine willingness to lean into change is what separates the firms helping clients navigate disruption from the ones still trying to plan their way around it.

“We want to be the disruptor,” said Walk. “We want to lean into disruption.”

Join the IDC Quanta Waitlist

Christina Cardoza - Content Marketing Manager - IDC

LinkedIn
Christina Cardoza is a Content Marketing Manager at IDC, where she specializes in brand content and social media strategy. With a background in journalism and editorial leadership, she has a proven ability to transform complex technology topics into clear, actionable insights.

2025年，智算云（AI IaaS）市场进入新一轮重构周期。行业演进的核心驱动力，已从单纯的模型训练需求，转向AI应用场景的快速扩张、底层技术瓶颈的持续突破，以及推理时代对算力组织方式的重新定义。训练、推理等工作负载与算力、网络、电力等基础设施开始深度耦合，推动AI IaaS从”GPU资源供给”逐步进化为涵盖算力底座、调度平台、模型服务与效能优化的全栈能力平台。

IDC认为，AI IaaS市场的竞争逻辑已发生根本性跃迁—— 从”资源供给”的单一维度比拼，转向以”算力底座、平台调度、模型服务与国产化生态”为核心的全栈能力竞争。这一转变，将决定未来三年市场格局的最终走向。

IDC最新发布的《中国智算云服务市场（2025下半年）跟踪》数据显示，2025下半年AI IaaS市场同比增速高达132.1%，市场规模达288.0亿元，2025全年市场规模达486.7亿元。（IDC定义下的AI IaaS覆盖GPU实例、AI专属集群以及相关AI存储服务，不含AI PaaS/SaaS层收入。）

算力资源成为AI产业核心壁垒

AI产业的竞争维度正在发生变化。过去行业更关注模型能力与产品体验，而当前，决定企业竞争上限的关键变量，正转向大规模算力资源的获取能力。对于头部基模厂商而言，pre-train、post-train、fine-tune、inference以及Agent工作流的持续扩展，都意味着对稳定智算资源的长期占用。高估值AI企业若想维持模型迭代速度与商业化增长，必须持续获得大规模、高稳定性的训练与推理资源支持。这一趋势也意味着，AI行业的竞争开始向更底层的物理资源延伸，包括土地、电力、数据中心以及先进芯片供应链。能够掌控这些资源的企业，将更容易建立长期竞争优势。除超大规模云服务商外，AI公司通常难以独立建设同等级基础设施，因此通过长期合作、预定集群乃至联合建设智算中心，正在成为行业主流路径。今年5月，Anthropic与xAI围绕算力资源展开合作，也进一步体现出“算力资源战略化”已经成为全球AI产业的重要趋势。

算力涨价潮背后的供需结构性矛盾

自2026年以来，AI算力产品价格持续上涨，反映出供需两端的多重刚性约束。需求端，AI模型规模和应用场景持续扩张，推理和训练算力需求同步激增。供给端，HBM高带宽存储、先进封装产能、电力基础设施等成为短板，算力边际增量有限，导致全球范围内算力涨价。国内市场上，今年3月，阿里云、腾讯云等头部厂商多轮上调AI算力及相关产品价格，反映出供需失衡的现实。涨价不仅体现在GPU集群租赁，还波及API调用、存储、网络等配套服务。IDC预计，短期内算力涨价压力仍将存在，推动企业优化算力采购策略，云服务商则需提升资源调度和能效管理能力以缓解供需压力。

推理时代到来，AI基础设施形态正在重塑

AI工作负载正在从“训练主导”逐渐转向“推理主导”。随着智能助手、AI Agent和自动化工作流逐步进入企业日常运营，AI调用已不再是单次问答，而是覆盖规划、检索、工具调用、代码执行、验证反馈等完整链路。一次任务往往对应数十甚至上百次模型推理请求，推理负载开始呈现高并发、低延迟、持续在线的新特征。

相比训练场景，推理更加关注：

  • 低延迟响应能力
  • 弹性扩缩容能力
  • 推理成本优化
  • 多模型混部调度能力
  • 边缘与中心协同部署能力

IDC数据显示， 2025年推理算力支出将首次接近训练算力支出。IDC认为，推理正在成为AI IaaS市场最重要的新增需求来源。

市场格局演化：“双阵营”结构形成

经过近一年的市场调整，中国AI IaaS市场逐渐形成“云服务商+运营商”双阵营竞争格局。云厂商延续了传统云计算时代在资源调度、产品体系和生态能力上的优势，并进一步向模型、推理服务与Token计费等AI链路延伸。

其中，阿里云依托灵骏智能计算集群，在大规模集群调度、异构算力支持、网络架构以及推理优化方面持续投入；百度智能云通过“芯云模体”体系，强化训练与推理一体化能力，在自动驾驶、金融、智能制造等行业推进落地；火山引擎则更多依托内部大规模推理场景积累，在推理成本优化与资源利用率方面建立差异化优势。运营商阵营则主要依托政企市场基础与国家战略资源，强化“算网融合”能力。其中，中国电信通过“息壤”平台推进全国算力调度；中国移动则通过“算力新动能行动计划”等持续强化算网协同能力。整体来看，市场集中度仍在提升，头部厂商在资金、资源获取和基础设施运营上的优势进一步扩大。

行业应用驱动市场结构重塑

当前，互联网与基模公司仍是GPU集群的主要采购方，但行业需求结构正在发生变化。互联网行业的音视频、电商、在线教育、社区平台等场景，正在持续扩大内容生成与推荐模型的推理需求；而基模公司则仍以训练和模型服务为主要资源消耗方向。与此同时，汽车自动驾驶场景和金融行业风控与智能投顾等场景也因其持续且庞大的计算需求，成为算力消费的重要力量。

国产化生态深化，供应链安全与自主可控成行业共识

外部供应链约束正在加速国产算力生态建设。过去，国产AI芯片主要受限于训练性能、生态兼容性以及集群稳定性，因此更多应用于边缘或特定行业场景。但随着推理需求快速增长，市场对“极致训练性能”的依赖有所降低，国产芯片在推理场景中的可用性明显提升。头部互联网公司、自动驾驶厂商等行业已在文本生成、图像识别、自驾仿真等多个场景中采用华为昇腾、百度昆仑芯、平头哥真武等国产AI芯片。供应链安全与自主可控，已经从政策导向逐步演变为行业共识。

IDC建议：AI IaaS竞争将进入“全栈能力”阶段

未来AI IaaS市场的竞争重点，将从单纯算力供给转向全栈能力竞争。

对于云服务商而言，需要持续强化异构算力布局、推理基础设施建设以及国产生态适配能力，同时围绕重点行业推出场景化解决方案，提升客户长期粘性。未来，API调用、Token计费、推理优化服务等新模式，预计将成为重要增长方向。

对于企业用户而言，则需要更加关注算力供给稳定性、推理性能、资源弹性以及整体成本效率。在推理需求快速增长背景下，按需租赁、弹性扩容与混合部署模式，有望在部分场景下逐步替代传统重资产采购模式，成为AI基础设施建设的新常态。

IDC 相关研究

围绕 AI 时代智算云与AI IaaS相关研究内容、技术能力与市场格局，IDC 将持续开展系统性研究，包括但不限于：

  • 《IDC Market Presentation：AI云存储市场进展及主流产品竞争力分析，2026》（即将发布）
  • 《IDC Market Presentation：中国企业智算云服务采购偏好调研，2026》（即将发布）

如需进一步了解 AI 时代智算云与AI IaaS相关研究内容，或咨询 IDC 在云计算、AI 基础设施及数字化转型领域的其他研究成果，欢迎与我们保持联系。

Qijin Chen

Qijin Chen - Senior Research Analyst

Analyst Profile
Chen Qijin currently works as a Senior Research Analyst for IDC China Enterprise Research. He specializes in analyzing and researching IT services, cloud professional/managed services, and DEOT services, providing industry analysis, forecasts, and market insights in related areas. In addition,…

2026年5月20日，全球大模型市场迎来“超级发布日”：谷歌在I/O大会连发Gemini 3.5 Flash、Gemini Omni两大新模型，阿里同步推出Qwen3.7-Max/Plus系列。中美巨头同周期密集迭代，直指Agent化、全模态、普惠高效三大核心方向，标志着大模型竞争从“参数竞赛”全面转向“能力落地+生态博弈”，全球大模型市场格局加速重构。

核心观点：技术路线齐向Agent，能力侧重各不同

谷歌主打全模态+极速性价比，锚定消费者Agent生态：谷歌推出多款 Gemini 系列模型， Gemini 3.5 Flash 速度更快、成本更低，综合能力升级，已是多款主流应用默认模型且用户体量庞大。同时谷歌依托专属平台与私人智能体布局核心战略，推动 AI 转变为全天候自主助手，深耕多类实用应用场景。全能模型 Gemini Omni 可融合多类媒介内容，具备多模态实力，还支持对话式视频编辑。谷歌未来战略以Antigravity 2.0平台+Gemini Spark私人智能体为核心，推动AI从 “工具” 进化为7×24小时自主执行的个人助手，全面渗透搜索、电商、创意场景。

阿里短短几月模型迭代至Qwen3.7，聚焦超长程Agent+部分开源：阿里 Qwen3.7 系列实力出众，Qwen3.7-Max面向智能体设计，在编程、推理等方面进行了能力更新，具备任务目标对齐能力，拥有 35 小时超长程自主执行能力，全新升级的Qwen3.7-Max即将通过阿里云百炼平台正式上线，全面对齐OpenAI与Anthropic API协议,并与Claude Code、OpenClaw及Qwen Code等主流智能体框架实现即插即用的无缝集成。视觉版 Qwen3.7-Plus 提升阿里视觉领域能力，打破海外多模态技术壁垒。此次发布是阿里云 “芯片-云-模型-推理” 全栈Agent化升级的一部分，该系列坚持部分开源的发展思路，着力搭建普惠型全球 AI 基础设施。

全球大模型迭代三大核心方向:可执行Agent、全模态融合、生态路线分化

从推理增强迈向 Agent 原生，行业竞争聚焦落地执行:  除了5月20日密集发布的模型，百度文心5.1也在本月稍早升级了智能体、知识、推理及深度搜索等方面的能力。可以看出Agent能力提升成为行业头部企业的共识，大模型发展下一阶段核心是智能体式思考，不再局限于基础逻辑思考。全球大模型企业将锚定智能体时代发展方向，优化模型长任务处理与工具调用能力，搭建平台助力智能体系统搭建，补齐长任务运行短板。整体行业趋势下，大模型正从单纯对话工具升级为自主生产力载体，评判标准也从精准度转为任务完成效率、运行稳定性等实战指标。

进阶全模态深度融合，角逐多模态技术高地：当前大模型赛道竞争正式进入全模态融合阶段，该能力已然成为旗舰模型必备配置。头部厂商打通文本、图像、音视频、3D 等多元内容交互链路，拓展创意生成新场景，加强视觉识别、逻辑推理与代码等能力，行业竞争重心也从单一能力比拼，转变为全模态协同能力与实际应用场景的全面比拼。

行业路线走向分化，普惠高效成为发展主流：大模型产业商业模式逐步分化，形成闭源商用与开源普惠或两者兼有的发展路径。谷歌依托闭源路线打造高性价比模型，凭借高速低成本优势抢占大众与企业市场，依托庞大用户体量构筑商业生态；阿里新模型坚守部分开源路线，降低开发使用门槛，汇聚全球开发者共建开放 AI 生态。两大阵营殊途同归，均聚焦能效比提升，摒弃盲目堆砌参数，以低成本轻量化部署作为核心发展优势。

四大路径助力产业破局：产品化、开源、生态和企业级特性

面对近期基础大模型带来的技术变革与市场机遇，IDC 给出如下建议助力中国大模型产业破局：

产品组合的广度与复杂性平衡：基础大模型产品组合的广度能带来优势，全球市场的头部厂商构建了高端推理，大规模部署， API 密集型长上下文场景，Agent场景，代码/软件工程场景，企业知识库/应用集成场景，前沿模型等专有模型。中国市场技术供应商可以考虑针对不同场景模型建立深度优化能力，而不是在所有场景中采取同质化的竞争策略；产品组合的广度也可能增加复杂性，可能增加客户在选择、集成、治理和总拥有成本（TCO）方面的复杂性。

开源与商业化策略：在大模型行业提速发展背景下，国内厂商持续优化开源与商业化布局。依托开源优势契合合规行业需求，兼顾成本、性能与市场需求，灵活平衡技术开放与商业落地，探索适配本土市场的可持续发展经营模式。开源模型采用开放且透明的设计，因此对那些要求模型具备高度透明度和合规性的受监管行业极具吸引力。对于将可解释性、可审计性和模型溯源列为采购要求的行业而言，能够检查模型权重、审核训练数据文档，并在无需依赖供应商的情况下修改模型行为，这本身就是一种竞争优势；另外，用户也在模型质量/推理深度、性能与预算之间寻求平衡， 中国技术供应商在以低成本触达用户的同时，还必须平衡模型质量/推理深度、速度与商业成果。

生态为王：中国大模型厂商需要与自有产品生态深度集成，并和全球广泛的上下游合作伙伴、系统集成商网络建立合作，并对全球范围内的开放协议积极加持；市场需广泛覆盖个人用户、开发者及企业级用户。

企业级特性纵深发展：在垂直领域，仍需配套企业级软件组合，通过生态合作伙伴网络加速行业知识的积累以及定制化解决方案的交付周期， 并配合专业服务。

IDC中国研究经理程荫表示，当前全球大模型正式迈入 Agent 迭代高速发展阶段，全球和中国市场技术供应商相继推出全新系列模型，开启行业新一轮竞争, IDC建议中国技术供应商从差异化产品、开源和商业化策略布局、全域生态搭建，到深耕企业级服务，多维度统筹发展，推动行业稳健提质、落地提速。

IDC在大模型与生成式AI领域有深入丰富的研究，欢迎广大技术供应商、行业用户、投资机构垂询。

请点击此处与我们联系。

Anne Cheng

Anne Cheng - Research Manager

Analyst Profile
Anne Cheng is a research manager in IDC China whose research focuses on the AI and big data markets. She collaborates with IDC's regional and global consulting teams and is involved in the business development of related markets. Prior to…