Rapidus Builds 17-Company Design Ecosystem for Its 2nm Foundry

Rapidus announced on October 5 that it is establishing Rapidus CORE, short for Collaborative Open Rapid Ecosystem, as a framework intended to connect semiconductor research, design resources, EDA and IP with its manufacturing operation. The first initiative is Design Solution Associates, a group of 17 companies that Rapidus says will help prospective customers with early-stage chip design.

The initial group includes Cadence Design Systems and Synopsys alongside AION Silicon, Dai Nippon Printing, GreatAsic Technology, HCL Technologies, Infinecs Systems, Infosys, MavissDesign, MosChip Technologies, NSW, Oppstar Technology, Quest Global Services, SkyeChip, TOPPAN, Toshiba Information Systems and Wipro. Rapidus says later CORE expansion will cover additional EDA, IP and other categories.

A leading-edge process is not useful to customers by itself. A chip company also needs design tools, verified IP, engineering expertise, physical-design flows and manufacturing support before a design can move toward tape-out and volume production. Rapidus is therefore trying to build the surrounding design infrastructure while it develops its manufacturing process.

The distinction matters when judging the announcement. The 17-company group expands Rapidus's customer-support ecosystem; it is not evidence that 17 customers have committed production orders, nor does it establish production yield, wafer pricing or commercial volume for the 2nm process.

Reuters reported on October 6 that Rapidus's state-backed advanced-chip effort is receiving about $15 billion in government support and is targeting 2nm mass production in late fiscal 2027. The report also highlights the central commercial challenge: attracting enough customers while competing with established advanced-node foundries including TSMC, Samsung and Intel.

That makes CORE relevant beyond the partner count. A new foundry must reduce the practical cost and risk of moving a customer's design onto an unfamiliar process. A broader design ecosystem can help with that transition, but successful high-yield manufacturing and committed volume business remain separate milestones that the October announcement does not prove.

Rapidus is part of Japan's effort to rebuild leading-edge semiconductor manufacturing capacity. If its 2nm program reaches commercial scale, it could add another advanced-node manufacturing option to a market currently concentrated among a small number of foundries. For PC and AI hardware, however, any downstream effect depends on actual customer designs, yields, capacity and production timing rather than the existence of the ecosystem alone.

The useful near-term signal is narrower: Rapidus is moving beyond process-development announcements and building the design-support layer needed to pursue commercial customers. The next evidence to watch is concrete customer engagement, validated design flows and IP, process qualification, yield progress and ultimately volume-production commitments.