English 
Rapidus has announced the development of a new suite of AI-powered semiconductor design tools aimed at dramatically improving design efficiency for next-generation chips, including those built on 2 nanometer (nm) processes. The announcement was made at SEMICON Japan 2025, where the company outlined its strategy to integrate artificial intelligence into the semiconductor design workflow as part of its broader Rapid and Unified Manufacturing Service (RUMS) initiative.
This new class of tools, collectively rebranded under the Rapidus AI-Agentic Design Solution (Raads), is scheduled for phased release beginning in 2026. Designed to work alongside traditional electronic design automation (EDA) tools, Raads leverages large language models (LLMs) and machine learning to help designers generate optimized hardware designs faster and with greater accuracy.
What’s New: Raads AI Design Tool Suite
Rapidus’ announcement highlights multiple tools that will soon become part of Raads:
- Raads Generator
An AI-driven EDA tool that takes natural language or specification inputs and automatically produces register-transfer level (RTL) design data tuned for Rapidus’ 2 nm manufacturing process. This accelerates early stage design workflows and reduces repetitive manual coding.
- Raads Predictor
This tool analyzes RTL source code along with Synopsys Design Constraints (SDC) to estimate key performance metrics, such as power, performance, and area (PPA), early in the design phase. By providing early insight into manufacturability, it helps designers refine architectures before costly iterations.
こちらもお読みください： 日本の大手銀行が半導体新興企業ラピダスに130億米ドルの融資を準備 - 国内チップ産業への戦略的後押し
Future Extensions
Additional tools scheduled for release throughout 2026 include:
- Raads Navigator / Indicator, which aids designers in resolving design challenges using LLM intelligence.
- Raads Manager, a layout planning tool that speeds hierarchical design.
- Raads Optimizer, which applies machine learning to identify optimal design parameters for improved PPA outcomes.
According to industry sources, these AI tools can reduce design time by up to 50 % and cut design costs by roughly 30 % when used in conjunction with existing EDA environments.
Why This Matters: The Intersection of AI and Semiconductor Design
Semiconductor design is one of the most complex engineering challenges in the technology industry. Modern chips, especially those using cutting-edge nodes like 2 nm with gate-all-around (GAA) transistor structures, require precise coordination between functional design, verification, and physical implementation. Traditional EDA tools have advanced steadily, but they still demand extensive manual effort and deep specialist knowledge.
The integration of generative AI and machine learning into the design process represents a paradigm shift known as AI-driven design automation. This approach uses advanced models to assist or augment human designers by automating routine tasks, suggesting optimized configurations, and predicting performance tradeoffs.
ラピドス’ Raads solution reflects this trend by embedding intelligence into every major step of the design workflow. Instead of writing verbose RTL manually and iterating through multiple cycles of simulation and tuning, engineers can leverage AI to automatically generate initial design structures and evaluate performance tradeoffs, ultimately shortening development cycles and increasing productivity.
日本のハイテク産業への影響
Boosting Domestic Semiconductor Competitiveness
Japan has been seeking to re-establish itself as a leader in global chip production, particularly for advanced logic and specialized semiconductors. Rapidus, founded in 2022 with backing from major corporations and government support, is a central part of that effort. The introduction of AI design tools reinforces Japan’s commitment to creating a complete domestic semiconductor ecosystem that includes design, manufacturing, and packaging.
By enabling designers to iterate faster and explore more architectural alternatives with AI assistance, Rapidus strengthens its ability to compete with global foundries and fabless partners. These tools also help lower barriers for smaller design firms that may lack deep expertise in ultra-scaled process technologies.
Accelerating Innovation Across the Supply Chain
AI design automation not only benefits Rapidus but also stimulates innovation across Japan’s semiconductor supply chain. Electronic design automation vendors, IP providers, and system integrators are likely to evolve their offerings in tandem with Raads, creating a richer ecosystem of AI-supported design solutions. Partnerships similar to Rapidus’ collaborations with tools and EDA vendors can accelerate joint progress and broaden access to advanced design methods.
Moreover, by decreasing cycle times and supporting design for manufacturability, AI tools can improve yield outcomes and reduce development costs, essential factors in the economics of advanced semiconductor manufacturing.
Effects on Businesses Operating in This Sector
Engineers and Design Teams
For system-on-chip (SoC) designers and engineering teams, Raads’ tools offer tangible improvements in productivity. Tasks that previously required manual scripting, constraint tuning, and iterative optimization can now be partially automated, enabling engineers to focus on higher-value innovation rather than routine tasks.
This could help attract and retain talent in Japan’s semiconductor sector, where competition for skilled engineers is intense. Providing cutting-edge AI design tools also supports upskilling efforts and makes Japan a more attractive place for research and development.
Fabless and IDM Companies
Japanese fabless firms and integrated device manufacturers (IDMs) stand to gain from better integration between design tools and manufacturing services. With AI-generated RTL optimized for Rapidus’ 2 nm process and early PPA feedback through Raads Predictor, companies can take a more co-optimized approach between design and fabrication. This reduces redesign loops and improves time-to-market for complex chips.
Broader Tech Industry Growth
AI-augmented semiconductor design fits into big global trends.
これには以下が含まれます：
- AI hardware acceleration
- エッジコンピューティング
- New mobile applications
- Automotive applications
Improved design speed and accuracy will help Japanese companies create specialized chips. These chips will help with AI tasks, sensor fusion, and processing that saves energy. These areas are key in the global tech race.
Conclusion: A Strategic Step Toward AI-Driven Semiconductor Future
Rapidus has launched the Raads AI design tool suite. This marks a key step in using artificial intelligence for semiconductor development. The company is using advanced generative models and automated design flows. This approach improves its internal processes and shows how AI can change hardware innovation.
This initiative boosts Japan’s tech industry. It boosts the country’s role in advanced semiconductor ecosystems. It also shows a move toward smarter and more collaborative design methods. AI design tools help engineers, businesses, and the supply chain innovate faster. They also lower costs and support global competitiveness in the coming decade.
日本語