Listen to this Post
Introduction
Japan’s Rapidus Corporation is making significant strides in the domestic production of cutting-edge semiconductors, aiming to rival global leaders in AI chip manufacturing. At the recent Semicon Japan 2025 exhibition in Tokyo, Rapidus unveiled a key prototype showcasing its latest innovation: a high-efficiency semiconductor substrate capable of hosting advanced AI chips. The company is signaling a strong intent to collaborate with NVIDIA, reflecting Japan’s broader push to localize semiconductor production and reduce reliance on foreign supply chains.
Rapidus Achieves Key Milestone in Semiconductor Substrate Development
On December 19, Rapidus CEO Junyoshi Koike announced that the company had completed the critical components of its semiconductor substrate designed for AI chips. The prototype, revealed at the Tokyo Big Sight exhibition center, incorporates an advanced “RDL interposer” wiring layer on a 600mm-square glass panel, designed to efficiently connect multiple AI chips. This design aims to maximize chip density while reducing production costs. Koike expressed a strong interest in hosting NVIDIA chips on this substrate, underscoring Rapidus’s ambition to become a contract manufacturer for major AI semiconductor companies.
Innovative Approach with Larger Panels
Typically, interposers are produced from 300mm round wafers. Rapidus, however, uses 600mm square panels, allowing up to 49 interposers per panel—a tenfold increase compared to traditional wafers. This approach not only reduces manufacturing costs but also supports larger, high-performance AI devices, which are steadily increasing in size due to evolving computational demands. Koike emphasized that the substrate’s dimensions exceed industry standards by 1.3 to 2 times, preparing the company for the next generation of AI semiconductors.
Showcasing Efficiency and Ambition
At the exhibition, Koike presented the panel in a manner reminiscent of NVIDIA CEO Jensen Huang, demonstrating Rapidus’s aspiration to become a manufacturing partner for global AI chip leaders. He humorously noted performing 100 push-ups daily to lift the panels, highlighting both the physical and symbolic effort behind Rapidus’s technological achievements.
Strategic Domestic Production for AI Chips
Founded by a consortium of Japanese companies including Toyota, NTT, and Sony Group, Rapidus is part of Japan’s strategic effort to produce next-generation semiconductors domestically. The company aims to establish mass production capabilities by 2028, with government support through subsidies and incentives to accelerate technological development. This effort positions Japan as a competitive player in AI hardware, particularly in the high-performance computing and AI chip sectors.
What Undercode Say:
Rapidus’s unveiling of its 600mm RDL interposer panel marks a pivotal step toward Japan’s semiconductor self-reliance. By departing from the conventional 300mm wafer approach, the company significantly increases yield and reduces cost per unit, which is critical in the high-margin, high-performance AI chip market. The ambition to host NVIDIA chips is not merely symbolic—it signals a strategic positioning to capture international partnerships while nurturing domestic production capabilities.
Technically, increasing interposer size by 1.3–2 times addresses the growing physical footprint of modern AI devices, allowing more chips per unit while maintaining efficient thermal management and connectivity. This scalability could give Rapidus a unique competitive advantage, as semiconductor performance increasingly depends on integrating multiple chips with high-speed interconnects.
From a geopolitical perspective, Rapidus’s initiative aligns with Japan’s national priority to reduce dependence on foreign semiconductor supply, particularly amid global chip shortages and increasing US-China tensions over technology. Domestic production backed by industry giants and government subsidies creates a robust ecosystem that can accelerate R&D, standardize manufacturing, and attract international collaborators.
Economically, the use of 600mm panels and mass-produced interposers can drive down production costs significantly. This may make Japan a cost-effective alternative to traditional semiconductor hubs, particularly for AI chips where unit cost and yield optimization directly influence adoption in supercomputers, autonomous vehicles, and AI-driven data centers.
Furthermore, Rapidus’s focus on contract manufacturing for established players like NVIDIA opens pathways for technology transfer, knowledge sharing, and international credibility. Successfully hosting foreign-designed chips domestically could create a new model for East-West semiconductor collaboration, blending Japanese manufacturing precision with global chip design leadership.
Long-term, the company’s roadmap toward 2028 mass production reflects careful planning around process maturity, scaling challenges, and supply chain optimization. By emphasizing interposer design and panel size innovation early, Rapidus positions itself to handle the growing complexity of AI semiconductor architectures without significant redesign, ensuring a faster go-to-market timeline.
Fact Checker Results:
✅ Rapidus unveiled a 600mm RDL interposer panel for AI chips at Semicon Japan 2025.
✅ The interposer size is 1.3–2 times larger than competitors’ standards, designed to reduce production costs.
❌ Claims about NVIDIA chip installation are aspirational, not yet confirmed by NVIDIA.
Prediction:
📊 Rapidus is likely to become a leading domestic AI chip manufacturer by 2028, with its innovative 600mm panels attracting international partnerships. The company may drive cost reductions in AI chip production and increase Japan’s competitiveness in the global semiconductor market. Rapidus’s collaboration with NVIDIA or other AI chip makers could also accelerate the adoption of Japanese-manufactured semiconductors in high-performance computing and AI-driven industries.
▶️ Related Video (90% Match):
🕵️📝✔️Let’s dive deep and fact‑check.
References:
Reported By: xtechnikkeicom_de31e7d5fbba9c4b9eb7d2c9
Extra Source Hub (Possible Sources for article):
https://www.facebook.com
Wikipedia
OpenAi & Undercode AI
Image Source:
Unsplash
Undercode AI DI v2
Bing
🔐JOIN OUR CYBER WORLD [ CVE News • HackMonitor • UndercodeNews ]
📢 Follow UndercodeNews & Stay Tuned:
𝕏 formerly Twitter 🐦 | @ Threads | 🔗 Linkedin | 🦋BlueSky | 🐘Mastodon
