East Asian Technology Intelligence
Japan & China technology, translated and contextualized for Western readers
富士通 / 富士通株式会社
A major Japanese enterprise IT-services and systems-integration provider, embedding AI into legacy, public-sector, and mission-critical operations through partnerships with global technology and cloud leaders.
Fujitsu is a large Japanese enterprise-technology provider whose AI relevance comes mainly from integrating third-party compute and AI software into complex customer environments. Its installed base and systems-integration capability support this role, while its lack of disclosed AI revenue and dependence on external GPU and cloud ecosystems constrain the investment case.
Fujitsu is primarily an enterprise IT-services and systems-integration company, not an AI accelerator, memory, foundry, or advanced-packaging supplier. Its direct AI offerings include Fujitsu Kozuchi, a platform for generative AI, agentic AI, and Composite AI; the Takane large language model; AI workflow orchestration; the planned Arm-based FUJITSU-MONAKA processor; 1FINITY optical and network products; and industry-specific integration services delivered through Fujitsu Uvance.
Service Solutions is the central business, generating 66.1% of FY2025 consolidated revenue. Hardware Solutions, which includes servers, storage, mobile base-station systems, and optical transmission systems, accounts for 26.6%. Ubiquitous Solutions, which includes PCs, makes up the remainder. The company's defensible position lies in enterprise integration, modernization of complex systems, Japanese-language and regulated-industry delivery, operational technology knowledge, and multi-decade customer deployments. It operates as a critical partner for Japanese government, financial-services, and telecommunications clients.
AI demand reaches Fujitsu revenue indirectly through Service Solutions. The company sells consulting, modernization, managed services, Uvance solutions, AI platform implementation, and infrastructure integration. It does not report a separate AI segment or disclose that a stated percentage of FY2025 revenue comes from AI. Service Solutions grew while Hardware Solutions declined, indicating that AI exposure currently has more to do with software, services, and customer transformation work than with standalone semiconductor sales.
The most concrete current infrastructure proposition is Fujitsu's collaboration with NVIDIA. The companies plan an AI computing platform that connects the future FUJITSU-MONAKA CPU series with NVIDIA GPUs through NVIDIA NVLink Fusion, with software combining Fujitsu's Arm optimization work and NVIDIA CUDA. Fujitsu also plans industry AI-agent software using Kozuchi, NVIDIA Dynamo, NVIDIA NeMo, and NVIDIA NIM microservices.
Fujitsu matters most in Enterprise Applications and Cloud & AI Platforms because its principal AI route to market is integration of AI into existing Japanese and multinational enterprise systems, rather than the sale of AI chips or cloud capacity.
Fujitsu's defensible position is in enterprise integration, modernization of complex systems, Japanese-language and regulated-industry delivery, operational technology knowledge, public-sector relationships, high-performance computing experience, and multi-decade customer deployments. Its competitive advantage does not rest on being a leading-edge semiconductor manufacturer or an owner of hyperscale cloud capacity.
The FUJITSU-MONAKA processor is the company's main bid to regain technical relevance in infrastructure. It is an Arm-based processor planned for release in 2027. The NVIDIA partnership improves its relevance because it pairs that planned CPU with NVIDIA GPUs through NVLink Fusion and combines Fujitsu's Arm software work with CUDA. However, this remains a roadmap and collaboration, not disclosed production volume or commercial revenue. Against global IT-services competitors, Fujitsu is smaller globally but more deeply embedded in some Japanese legacy environments.
Fujitsu depends on external accelerators, cloud platforms, networking software, semiconductor manufacturing, and ecosystem software. NVIDIA provides GPU, CUDA, NIM, NeMo, Dynamo, and NVLink Fusion components in the joint offering. Palantir supplies AIP software under a licensing agreement. Arrcus supplies routing and switching software for 1FINITY network offerings. Microsoft remains a cloud and AI partner.
On the customer side, Fujitsu publicly identifies Sony Bank, Rohto Pharmaceutical, and Science Tokyo as partners or customers in specific programs. It does not publicly disclose revenue by named AI customer, customer-level AI revenue, or an AI-revenue share. Its exit from Shinko Electric removes a potentially important direct connection to advanced semiconductor packaging, leaving more of its AI case dependent on integration rather than component supply.
Fujitsu's most visible AI-infrastructure roadmap relies on NVIDIA GPUs, CUDA, NIM, NeMo, Dynamo, and NVLink Fusion, so NVIDIA supply, licensing terms, and product roadmaps constrain Fujitsu's offering. Products built around advanced NVIDIA GPUs and related software can be subject to U.S. export controls, affecting Fujitsu's ability to deploy AI infrastructure in some markets and for some end users.
Fujitsu's core strengths are closely tied to Japan's enterprise, public-sector, and financial-services systems, creating sensitivity to domestic IT spending cycles and modernization budgets. Its servers, networks, and future MONAKA-based infrastructure depend on third-party chips, manufacturing capacity, and components that can be disrupted by regional conflict, logistics shocks, or supply constraints.
| Risk | Severity | Why it matters |
|---|---|---|
| Dependence on NVIDIA's AI ecosystem | High | NVIDIA supply, licensing terms, and product roadmaps constrain Fujitsu's offering. |
| U.S. export-control exposure | High | Products built around advanced NVIDIA GPUs can be subject to U.S. export controls. |
| Services-business execution | High | Project delays, labor constraints, or unprofitable contracts can materially affect earnings. |
| AI revenue remains undisclosed | Medium | Difficult to determine whether AI product activity is material to group financial performance. |
| Japanese-market concentration | Medium | Sensitivity to domestic IT spending cycles and modernization budgets. |
| Legacy-system liability | Medium | Implementation risk, cybersecurity incidents, system outages, and costly delivery failures. |
| Hardware decline | Medium | Hardware Solutions revenue fell 11.0% in FY2025. |
| Semiconductor and component sourcing | Medium | Dependence on third-party chips and manufacturing capacity. |
Fujitsu is an independent, publicly listed Japanese company. It is not state-owned and is not identified as controlled by a founding family, chaebol, industrial parent, or Japanese government entity. The shareholder base is dispersed, with The Master Trust Bank of Japan, Ltd. held for trust as a principal shareholder in a custodial structure.
Fujitsu uses a board structure with an Audit and Supervisory Committee, including an Executive Nomination Committee and a Compensation Committee. The company operates under IFRS for consolidated reporting, having adopted it from fiscal 2014 to improve consistency and comparability in international capital markets.
Compiled with AI-assisted research from company filings, market data, and published reporting as of September 12, 2026, then reviewed by AsiaAI.FYI. Figures marked Estimate are not company-reported. Check primary filings before relying on any number.
Confidence: B. Core financial data, segment data, and capital-return authorization are supported by company reporting, while AI revenue, AI customer concentration, and FY2025 R&D expense remain undisclosed.
Main sources: Fujitsu FY2025 consolidated financial-results release; Fujitsu Integrated Report 2025; Fujitsu corporate facts page; Company press releases; Market data (Yahoo Finance, MarketWatch).
East Asian Technology Intelligence
Japan & China tech news — translated, contextualized, and delivered for Western readers.
Free. Unsubscribe anytime.