Samsung Electronics' chip unit is preparing to phase out the foreign analytics platform Spotfire, replacing it with an in-house AI system to cut license fees that have surged multiple times over the past year. Frontline teams have already begun building internal alternatives as the DS Division accelerates its AI-first manufacturing roadmap.
TL;DR
- Samsung's semiconductor division is replacing the foreign analytics tool Spotfire with in-house AI to reduce soaring license fees and strengthen data security.
- The move aligns with Samsung's broader AI transformation push, including a $73.5B investment in AI and semiconductor R&D through Q3 2025.
- Field concerns remain over data migration risks and accountability for AI-driven yield analysis errors.
What happened
Samsung Electronics' DS Division has launched an internal review to replace Spotfire, a widely used foreign analytics tool for semiconductor yield management, with proprietary AI systems. The decision follows a sharp increase in licensing fees, which have climbed several-fold compared to prior levels, according to semiconductor industry officials and Samsung employees.
Frontline teams are already developing internal alternatives to replace Spotfire's core functions, including real-time process data interpretation and defect root-cause analysis. The DS Division conducted a survey of working-level staff to assess feasibility, with plans to finalize decisions by next year.
The shift is part of Samsung's company-wide AI transformation (AX) initiative, which aims to embed agentic AI across R&D and manufacturing processes. At the Samsung AI Forum 2026, the company unveiled six task-specific agentic AI tools for semiconductor manufacturing, including models that analyze wafer maps and process histories to predict defects and recommend corrective actions.
Samsung's strategic partnership with Mistral AI is central to this effort, with plans to co-develop an on-premise AI model to ensure sensitive manufacturing data remains within internal networks, addressing security concerns tied to cloud-based analytics.
Why it matters
For AI developers and startup founders, Samsung's move underscores the growing demand for proprietary AI systems in semiconductor manufacturing, where data security and cost control are critical. Startups building on-premise AI tools for industrial applications may see new opportunities as incumbents seek alternatives to foreign software.
For investors, the shift highlights the strategic importance of AI infrastructure in hardware supply chains. Samsung's $73.5B investment in AI and semiconductor R&D through Q3 2025 signals confidence in internal AI capabilities, potentially reducing reliance on third-party tools and reshaping B2B software markets.
However, the transition poses risks for semiconductor manufacturers. Migrating vast volumes of process data and ensuring accuracy in AI-driven yield analysis remain unresolved challenges. Concerns about accountability for errors—whether borne by engineers or the AI system itself—could slow adoption if not addressed.
The broader implication is a potential shake-up in the B2B analytics software market, as major manufacturers prioritize internal AI solutions over licensed tools. This could pressure vendors like Tibco (Spotfire's owner) and accelerate consolidation in the industrial AI space.
Key facts
- Samsung's DS Division is evaluating replacing Spotfire with in-house AI due to license fees that have risen several-fold over the past year.
- Frontline teams are already developing internal analytics tools to replace Spotfire's functions, with plans to finalize decisions by next year.
- Samsung's AI transformation roadmap includes six agentic AI tools for semiconductor manufacturing, unveiled at the Samsung AI Forum 2026.
- Samsung invested "hundreds of billions of won" in Mistral AI and is co-developing an on-premise AI model to secure manufacturing data.
- Samsung's cumulative capital expenditure through Q3 2025 reached 56.944 trillion won ($42.5B), with 92.6% allocated to the DS Division.
- Total R&D spending through Q3 2025 was 41.984 trillion won ($31.4B), bringing combined AI and semiconductor investment to 98.424 trillion won ($73.5B).
- Samsung plans to complete AI training for all 2,300 employees by February 2027 as part of its "AI Great Transformation" initiative.
- A Samsung Electronics official confirmed the internal survey but stated that no final decision on Spotfire's use has been made.
Context
The semiconductor industry has long relied on third-party analytics tools like Spotfire for yield management, but rising costs and geopolitical tensions have intensified efforts to develop proprietary AI systems. Major manufacturers such as TSMC and Intel have also explored in-house AI solutions for similar reasons.
Samsung's push aligns with a broader trend in industrial AI, where companies seek to reduce dependence on foreign software amid supply chain and security concerns. This shift is expected to reshape the B2B analytics market, favoring vendors offering on-premise or customizable AI solutions.
The company's $73.5B investment in AI and semiconductor R&D underscores the scale of its commitment to internal AI capabilities, contrasting with its historical reliance on external tools and partnerships.
