rsscloudcomputing

AI Inference Infrastructure Optimization Demands Token Economics–Driven Decision-Making

Fri, 26 Jun 2026 04:00:00 GMT

This IDC Perspective shares an overview of how, as enterprises scale AI from experimentation to production, the explosive growth of inference workloads and agentic AI pipelines is making token cost management a board-level financial priority. Token economics (aka “Tokenomics”), the discipline of measuring, governing, and optimizing the full cost of AI token production and consumption, is emerging as the essential framework for every AI infrastructure, architectural, and purchasing decision.

IDC recommends that technology buyers build a full-stack token cost model capturing all infrastructure layers, including compute, storage, networking, power, cooling, and datacenter facilities, when comparing on-premises/dedicated versus cloud-based AI infrastructure options. “Enterprises need to reframe AI infrastructure decisions around token economics to accurately represent the true cost and benefits of AI infrastructure investments at scale,” explains Mary Johnston Turner, IDC research vice president, AI Infrastructure Strategies.

Databricks' Data + AI Summit 2026: Expanding into the Application Platform Market

Matthew Flug — Fri, 26 Jun 2026 04:00:00 GMT

At its annual Data + AI Summit, Databricks advanced a strategy to extend from the company's data management foundation into the application platform and agent development, deployment, and governance layers, anchored by Databricks Apps, Unity Catalog, Unity AI Gateway, Agent Bricks, and the unification of transactional and analytical data through Lake Transactional/Analytical Processing (LTAP). The announcements position the company to compete for agentic application workloads on the strength of governed access to enterprise data as it continues to mature its offerings relative to established application platform providers.

IDC Market Glance: Digital Twins in Regional and Local Government, 2Q26

Alison Brooks, Ph.D., Ruthbea Yesner — Fri, 26 Jun 2026 04:00:00 GMT

Digital twins are crossing into mainstream use in regional and local government. Globally, 42% of state and local government organizations report using digital twins today; another 25% are planning additional investment, and more than one-fifth plan to start within the next 12 to 18 months. This IDC Market Glance maps the vendor landscape across five segments: place-based digital twin platforms, industry-specific twins (covering critical infrastructure and utilities, public safety and emergency response, transportation and mobility, and environmental and climate monitoring), physical world data acquisition, digital twin enabling technologies, and building and facility twin platforms.

Three forces are reshaping this market. First, the cost and complexity of implementation have dropped materially: lower sensor costs, edge processing, and AI-native APIs have made real-time physical system modeling accessible to governments that could not have justified the investment five years ago. Second, partnerships among geospatial leaders, systems integrators, and hyperscalers are becoming the dominant delivery model, because no single vendor can cover the full stack from data acquisition through simulation and action. Closed, monolithic vendors risk being designed out of major government programs as a result. Third, the definition of what a twin can represent is broadening: governments are deploying policy twins, mission twins, and ocean environment twins alongside city-scale infrastructure models, and each new application type creates demand for new categories of vendor capability.

For government buyers, the decision-making calculus starts with mission outcome, rather than platform considerations. Investments anchored to a specific operational problem and funded to sustain the live data pipelines that keep the twin current are delivering lasting value. A model without a continuously updated sensor, camera, and operational system feeds reverts to a static visualization. For vendors, the market's direction toward open APIs and AI-native interoperability by default is now considered table stakes; requests for proposal (RFPs) are already reflecting this expectation. Policy and mission twins are on track to become the default decision-support layer for major public programs in the near term, making the current window a critical one for establishing platform and partnership position.

IDC Survey: AI on Cloud Infrastructure Adoption Trends in META, 2026 — Compute, Storage, Networking, and Silicon Choices

Jebin George — Fri, 26 Jun 2026 04:00:00 GMT

This IDC Survey examines how organizations across the Middle East, Türkiye, and Africa (META) are building and running AI in the public cloud. Based on IDC’s Cloud Pulse Survey, 1Q26, of 320 decision-makers in Saudi Arabia, the UAE, Türkiye, and South Africa, this report maps the AI infrastructure behind production AI and GenAI workloads across cloud AI spend, GPU and accelerator usage, compute platforms (VMs, containers, bare metal, serverless), storage, and networking for training and inference. It reveals where META differs: AMD’s Instinct MI300X leads as the most-used accelerator, AWS and Microsoft Azure co-lead, and silicon choices vary sharply by country. An essential read for cloud service providers, infrastructure vendors, and enterprise IT leaders, refer to the full report to benchmark your cloud AI strategy.

Market Share: Worldwide Public Cloud Platform as a Service Shares, 2025 — The Rise of Agentic AI

Adam Reeves — Fri, 26 Jun 2026 04:00:00 GMT

This IDC presentation reports on market share for the worldwide public cloud PaaS market. It includes reporting on the prior year, in addition to 2025, to provide context for how the market has evolved. It also provides an analysis of the market players and associated secondary markets.

"The top 3 vendors in the market for public cloud PaaS in 2025 were, in order, Microsoft, AWS, and Google Cloud, accounting for greater than 40.0% of the overall PaaS market. The AI-driven market shake-up taking place is demonstrated by OpenAI's continued ascent to now become the fourth-ranked vendor in the market with nearly 6.0% market share." — Adam Reeves, research director, Platform as a Service (PaaS) for Developers of Modern and Edge Applications

Worldwide Converged and Hyperconverged Systems Market Overview, 4Q25

Rajiv Ranjan — Fri, 26 Jun 2026 04:00:00 GMT

This IDC Market Presentation provides a quarterly (4Q25) and full-year (2025) update on the performance of the overall converged systems market and its three subsegments: hyperconverged systems, integrated platforms, and certified reference systems and integrated infrastructure. It also provides an overview of the market's performance over the past five quarters, broken down by subsegment, vendor, and region.

アクロニスが挑むバックアップベンダーからサイバープロテクション、総合プラットフォームベンダーへの転換

Mayumi Nishimura — Fri, 26 Jun 2026 04:00:00 GMT

本調査レポートは、2026年5月20日に開催されたAcronis Cyber Frame記者発表会（Acronis Cyber Day Japan 2026）への参加、およびアクロニス・ジャパンの代表取締役社長である川崎哲郎氏とアクロニスのプラットフォームマーケティング&エデュケーションシニアディレクターであるリックヘブリー氏による発表内容を基に作成している。同発表会においてAcronisは、Acronis Cyber Frame（HCI/IaaS）を中心に、Acronis MDR by Acronis TRU（マネージド脅威検出と対応）、Acronis GenAI Protection（生成AI利用ガバナンス）を含む一連の戦略製品を発表した。VMware代替需要の急増、SOC（Security Operations Center）スキルギャップ、AI利用リスクという3つの市場課題に対するAcronisの戦略的回答を考察し、MSPおよびIT意思決定者に対する示唆を提示する。

中国私有云CWPP市场份额，2025：私有云CWPP加速融合云原生技术，依托AI实现轻量化智能防护

James Wang, Joe Zhao — Fri, 26 Jun 2026 04:00:00 GMT

本次IDC研究主要阐述2025年中国私有云云工作负载保护平台（CWPP）市场的规模、厂商份额以及技术发展变化等内容。

“2025年私有云CWPP市场小幅收缩，头部格局稳固。AI深度赋能产品，智能研判、业务基线、多模态防护广泛应用，同时新型AI攻击加剧风险。行业走向平台化整合，租户隔离、防勒索、异构算力适配等技术需求持续上升。”——赵义，IDC中国安全和信任领域高级研究经理

Market Forecast：国内AI市場予測、2026年～2030年

Takuya Uemura — Thu, 25 Jun 2026 04:00:00 GMT

本調査レポートは、国内AI（Artificial Intelligence）市場について、2025年5月下旬までに収集した情報を基に定性的かつ定量的な分析結果をまとめ、2030年までの市場支出額を予測するものである。

国内AI市場は2025年に大幅な成長を記録し、AIエージェントの実ビジネス適用を起点に本格的な拡大局面に入ったとIDCではみている。予測期間を通じて高い成長が続き、特にソフトウェアが市場全体を大きく上回る成長を示す。なかでも、基盤モデルを含むAIプラットフォームは最高成長を記録し、AIエージェント時代のテクノロジー競争の主戦場となる。また、ハードウェアはセミソブリンAI投資が牽引し、サービスはPOC（Proof of Concept）から本番移行と業務変革支援を背景に着実に拡大する。「国内AI市場は、AIエージェントの業務利用と実用化を起点に、ソフトウェア／サービス／ハードウェアが相互に牽引し拡大する局面に入った。成長を主導するのはAIプラットフォームを含むソフトウェアであるが、その普及はインフラ投資や導入／業務変革を支えるサービス需要も同時に押し上げる。ITサプライヤーには、エコシステム全体で価値を提供する競争力が求められる」と、IDC Japan、AI and Automation のシニアリサーチマネージャーである植村卓弥は述べている。

Effective Agent Cost Management: A Practitioner's Framework for Planning, Sourcing, Implementation, and Operations

Wed, 24 Jun 2026 04:00:00 GMT

This IDC Perspective provides a comprehensive, practitioner-focused framework for effective AI agent cost management, emphasizing that every dollar of wasted spend is a governance failure. Covering planning, sourcing, implementation, and operations, it details actionable steps for budgeting, cost modeling, governance, and continuous optimization. The framework guides organizations to transform unpredictable AI costs into strategic, auditable investments by integrating financial discipline, technical controls, and cross-functional accountability throughout the AI agent adoption journey.

"Every dollar of wasted AI agent spend is a governance failure," according to Shari Lava, group vice president, AI at IDC. "Are you ready to turn unpredictable costs into strategic, auditable investments?"