Intellifusion Q3 Report: R&D Investment of 513 Million Yuan, Up 73.01% Year-on-Year, Self-Developed AI Chip Layout Accelerates

On October 31, Intellifusion (688343.SH) released its third-quarter 2026 report. The report shows that in the first three quarters the company achieved operating revenue of 944 million yuan, remaining broadly stable overall. R&D investment reached 513 million yuan, a year-on-year increase of 73.01%, and R&D investment as a proportion of operating revenue further increased compared with the same period last year.

As the artificial intelligence industry accelerates its development, AI is moving from breakthroughs in model capabilities toward large-scale application. AI Agents, video generation and various intelligent applications continue to emerge, driving growth in Token consumption and multimodal computing demand, and also placing higher demands on the performance, efficiency and cost of underlying computing power.

Against this backdrop, Intellifusion continues to increase its resource investment in self-developed AI chips. On the one hand, it strengthens key technology breakthroughs by bringing in high-level core R&D talent; on the other hand, it provides supply chain assurance for the subsequent launch of chip products by locking in advanced-process wafer foundry capacity in advance.

As AI applications continue to deepen, computing power demand is reflected not only in the model training stage but also throughout the continuous operation of various intelligent applications. In particular, emerging applications such as AI Agents and multimodal generation are driving demand for computing power toward higher frequency, larger scale and greater diversity.

According to disclosures in the financial report, one important reason for the growth in Intellifusion's R&D investment is that the company has continued to bring in high-level core R&D talent around key technology directions, with corresponding increases in related employee compensation and share-based payment expenses.

For the technology-intensive chip design industry, a high-level R&D team is an important foundation for continuously advancing technological innovation. AI chip R&D involves multiple technical links such as chip architecture design, algorithm optimization, basic software and toolchains, placing high demands on cross-disciplinary R&D capabilities.

The continued introduction of R&D talent and the accumulation of technical capabilities will provide support for product R&D, subsequent iterations and the evolution of the technology system.

Advance layout of advanced-process capacity to strengthen chip industrialization assurance

In this financial report, Intellifusion's layout in upstream wafer foundry capacity is also noteworthy.

According to disclosures in the financial report, affected by tightening supply of upstream advanced-process wafer foundry capacity, Intellifusion, in light of its business development needs, locked in wafer foundry capacity in advance through prepayments. This arrangement increased the company's cash paid for goods purchased and services received year on year, and placed a temporary occupation on operating cash flow during the reporting period.

From the perspective of the industrial chain, chip design companies need not only to complete design and R&D but also to coordinate multiple links such as wafer manufacturing, packaging and testing before product delivery can ultimately be achieved. Especially for chip products using advanced processes, the availability of wafer foundry capacity is directly related to subsequent production arrangements and product delivery capability.

As the scale of AI applications continues to expand, the importance of computing infrastructure construction and chip supply capability becomes further prominent. For chip companies, coordinating upstream capacity resources in advance is an important step in ensuring the connection between product R&D and production and advancing industrialization.

By locking in advanced-process capacity through prepayments, Intellifusion strengthens supply chain resource preparation while advancing chip R&D, creating conditions for subsequent product production and business expansion.

At present, the artificial intelligence industry is moving from technological breakthroughs toward large-scale application. From intelligent assistants to AI Agents capable of autonomously executing complex tasks, and from text generation to multimodal content creation such as images and videos, the boundaries of AI applications are constantly expanding, and Token consumption and related computing demand are growing accordingly.

As an enterprise that has long been deeply engaged in the AI inference field, Intellifusion continues to carry out technology R&D and product layout around self-developed chips, and takes key resources such as core R&D talent and advanced-process capacity as investment priorities, continuously stepping up core technology R&D and chip industrialization layout, so as to provide more efficient and economical computing power support for the growing computing power demand in the AI era.