In February 2025, the chip IP provider Arm announced the launch of its first self-developed chip, a news that has caused a strong shock in the global semiconductor industry. As a leading company in the chip architecture field, Arm is transitioning from a traditional IP licensing model to a strategy of self-designed and manufactured chips, which not only reshapes the competitive landscape of the industry but also initiates a profound wave of change in terms of technological autonomy and control.
Arm’s evolution of its business model is a strategy to gradually extend control over the chip design industry chain. In the first phase, Arm only provided instruction set architecture (ISA) licensing, allowing customers to design microarchitectures independently; in the second phase, Arm launched processor IP cores (such as the Cortex series), which customers could directly integrate and use; in the third phase, Arm introduced computing subsystems (CSS), providing pre-validated processor design solutions; and now entering the fourth phase—directly launching self-developed chips. This progressive vertical integration enhances Arm’s control over the industry chain while continuously compressing the “self-research” space for customers.
Of course, as we know, ARM is a company controlled by the fund of Japan’s richest man, Masayoshi Son, who has already promised to invest $1 trillion in front of Trump at the beginning of the year. The huge spending on AI data centers by major global tech giants makes ARM unwilling to settle for just earning a small amount of chip design licensing fees.
In the context of the escalating technological competition between China and the United States and the ongoing turmoil in the global supply chain, how domestic chip companies respond to the impact of Arm’s business model evolution and adjust and optimize their autonomous control strategies is crucial. This article will analyze the impact of Arm’s self-developed chips on the industry from multiple dimensions, including instruction sets, microarchitecture, toolchains, supply chains, industrial ecology, and industrial policies, and propose targeted and forward-looking recommendations for domestic chip companies.
1. The Multidimensional Impact of Arm’s Self-Developed Chips on the Industry
(1) Monopoly in the Instruction Set Field
Arm’s instruction set architecture has long dominated the mobile chip field due to its low power consumption and high performance. According to statistics, over 90% of smartphone chips worldwide are based on Arm’s instruction set design, giving Arm absolute authority in the instruction set ecosystem. The self-developed chip launched this time is based on the latest v9 architecture, which has significantly upgraded security and AI performance, further consolidating Arm’s technological advantage at the instruction set level.
For companies relying on Arm’s instruction set licensing, Arm’s transformation brings many potential risks. On one hand, to obtain the licensing for the v9 architecture, companies may need to pay higher licensing fees. For example, a well-known smartphone chip manufacturer previously paid about 15% – 20% of its chip R&D costs in licensing fees for Arm’s instruction set, and with the advancement of Arm’s self-developed chip strategy, it is expected that future licensing fees may rise to 30% or even higher. On the other hand, Arm may impose stricter restrictions on the use of the instruction set, including application scenarios and technology output. More seriously, in the context of a complex and changing international political situation, there is a risk of being cut off from instruction set licensing, which will directly threaten the survival and development of companies.
This potential monopoly threat not only deepens the industry’s dependence on Arm’s instruction set but also severely squeezes the development space for other instruction set architectures. Currently, apart from the two mainstream instruction sets, Arm and x86, the open-source RISC-V instruction set has developed rapidly but still holds less than 5% market share. Arm’s monopoly position makes it difficult for emerging instruction sets like RISC-V to gain sufficient resources and support, hindering the diversification of technology in the industry and adversely affecting the long-term healthy development of the chip industry.
(2) Intensified Competition in Microarchitecture Design and Ecological Restructuring
In the field of microarchitecture design, Arm has a deep technical accumulation and strong R&D capabilities. Its self-developed chip’s microarchitecture excels in performance and power optimization. For example, Arm’s latest CPU microarchitecture has improved performance by 20% – 30% at the same power consumption compared to the previous generation, while reducing power consumption by 15% – 20% at the same performance. This leading technological advantage poses a significant impact on traditional IC design companies like Broadcom and Marvell.
These companies have long relied on Arm’s IP cores for chip design. The launch of Arm’s self-developed chips not only exposes them to direct market competition from Arm but may also limit their microarchitecture design autonomy and flexibility due to Arm’s adjustments to its IP core strategy. For instance, Arm may prioritize applying the latest microarchitecture technology to its self-developed chips, while updates to the IP cores for licensed customers may lag behind, putting these companies’ products at a disadvantage in market competition.
Moreover, Arm’s self-developed chip strategy may accelerate the restructuring of the microarchitecture ecosystem in the industry. To reduce dependence on Arm and mitigate competitive risks, more and more customers are turning to self-developing microarchitectures or adopting open-source RISC-V architectures. For example, SiFive’s P870-D processor, developed based on the RISC-V architecture, has achieved performance comparable to some of Arm’s products, attracting attention and collaboration from many companies. This indicates that Arm’s transformation is driving the microarchitecture ecosystem in the industry from a singular to a diversified direction.
(3) The Trend of Toolchain Ecosystem Closure
The chip design toolchain is a key infrastructure for chip R&D, covering various aspects such as EDA software, compilers, and simulators. Currently, the global EDA market is mainly dominated by three American companies: Synopsys, Cadence, and Mentor Graphics, with a market share exceeding 70%. Although Arm is not a leader in the toolchain field, its transition to self-developed chips may strengthen its cooperation with toolchain vendors, promoting the formation of a toolchain ecosystem more favorable to its chip design.
For example, Arm may collaborate with EDA vendors to develop dedicated design tools for the v9 architecture, which may be more compatible in functionality and performance with Arm’s self-developed chips, while other companies using Arm architecture for chip design may not enjoy the same level of technical support and optimization services. This trend will lead to a gradual closure of the toolchain ecosystem, making companies using Arm architecture more limited in their toolchain choices, increasing the cost and difficulty of chip design. At the same time, a closed toolchain ecosystem will further consolidate Arm’s position in the entire chip design process, weakening the competitiveness of other companies and creating a situation where the strong get stronger.
(4) The Shock and Restructuring of the Supply Chain Landscape
Arm’s self-developed chips are manufactured by specialized foundries like TSMC, which further strengthens TSMC’s position in the high-end chip manufacturing field. TSMC, with its advanced process technologies such as 3nm and 2nm, can provide high-quality chip manufacturing services for Arm. Additionally, as Arm has established partnerships with tech giants like Meta, it is expected that a large number of orders will flow to TSMC in the future, changing the existing chip supply chain order distribution pattern.
For domestic chip companies, in the context of a tight global supply chain, Arm’s transformation undoubtedly exacerbates supply chain risks. On one hand, the limited capacity of foundries like TSMC means that the increase in orders for Arm’s self-developed chips will make it even more difficult for domestic companies to access advanced manufacturing processes. For example, SMIC currently has relatively insufficient capacity in advanced process technologies of 7nm and below, making it difficult to meet the growing demand from domestic companies. On the other hand, the supply of key raw materials also faces challenges. For instance, core materials such as photoresists and high-purity silicon wafers, which are essential for chip manufacturing, have long relied on imports, and the stability of supply is difficult to guarantee in the context of a complex international situation. Furthermore, the continuous escalation of the U.S. technological restrictions on China’s semiconductor industry further hinders domestic companies’ access to advanced equipment and technology, putting immense pressure on domestic chip companies in the supply chain.
(5) The Destruction and Differentiation of the Industrial Ecology
For a long time, Arm has built a relatively stable chip industry ecology as a neutral IP provider. In this ecology, Arm provides architecture licensing to numerous chip design companies, which develop a diverse range of chip products based on Arm’s architecture, forming a good situation of mutual cooperation and common development. However, the implementation of Arm’s self-developed chip strategy has transformed it from a builder of the ecosystem to a participant and competitor, which may disrupt the original balance of the industrial ecology.
On one hand, the competitive relationship between Arm and its customers significantly reduces the trust of partners in Arm. For example, companies like Broadcom may reduce cooperation with Arm and seek other architecture suppliers or increase investment in self-research, leading to differentiation in the industrial ecology. On the other hand, Arm’s extension downstream in the industry chain, leveraging its advantages in architecture, has squeezed the survival space of small and medium-sized chip design companies. For these small and medium-sized chip design companies, their relative weaknesses in technology, funding, and market make it difficult to compete with giants like Arm, potentially facing the risk of elimination, which is not conducive to the diversified development of the industrial ecology. Additionally, the launch of Arm’s self-developed chips may trigger price wars and market share competition within the industry, further disrupting the order of the industrial ecology.
(6) The Chain Reaction at the Industrial Policy Level
In the context of increasingly fierce global semiconductor industry competition, countries are introducing industrial policies to support the development of their domestic chip industries. Arm’s self-developed chip strategy may trigger a re-examination and adjustment of semiconductor industry policies in various countries.
For the United States, as the dominant player in the global semiconductor industry, it has always regarded semiconductor technology as an important means to curb competitors. Although Arm is headquartered in the UK, it also holds an important position in the U.S. semiconductor industry. The U.S. may further strengthen its control over companies like Arm, using them as tools to suppress the development of China’s chip industry. For example, through export controls and technology blockades, it may restrict Arm from providing technology and products to Chinese companies, intensifying the pressure on China’s semiconductor industry.
For regions like Europe and Japan, to ensure the safety and development of their own chip industries, they may increase support for domestic chip architecture and design companies. For instance, Europe has launched the “European Chips Act,” planning to invest 43 billion euros in semiconductor industry development, focusing on supporting local companies in technological innovation in chip architecture and design. Japan is also actively promoting semiconductor industry revitalization plans, strengthening cooperation with domestic companies to enhance the competitiveness of its local chip industry.
For China, Arm’s transformation undoubtedly exacerbates the pressure of external technological blockades, prompting the Chinese government to accelerate the improvement of the chip industry policy system and increase support for the autonomous and controllable chip industry. In recent years, the Chinese government has introduced a series of policy measures, such as establishing a national integrated circuit industry investment fund with a total scale exceeding one trillion yuan to support the technological R&D and industrial development of chip companies; formulating policies like the “National Automotive Chip Standard System Construction Guide” to guide the standardized development of the industry. In the future, the Chinese government may further increase policy support to promote the autonomous and controllable development of the chip industry.
2. In-Depth Adjustment of Domestic Chip Companies’ Autonomous Control Strategies
(1) Instruction Set: Building the Core Foundation of Independent Innovation
Increase Investment in Instruction Set R&D
Domestic companies should regard instruction set architecture R&D as a core strategy, drawing on the successful experience of Loongson Technology’s LoongArch instruction set. Loongson Technology designed the LoongArch instruction set from scratch and successfully broke free from dependence on Western instruction sets after years of R&D and accumulation. Domestic companies should form specialized R&D teams composed of computer architecture experts and chip design engineers, increase funding investment, and establish special R&D funds. At the same time, actively carry out industry-university-research cooperation, establishing joint laboratories with universities and research institutions to jointly tackle key technical challenges in instruction set design.
In the R&D process, it is essential to fully consider future technological development trends, focusing on the advancement, compatibility, and scalability of the instruction set. For example, with the rapid development of artificial intelligence and the Internet of Things, the instruction set should provide good support for AI computing and edge computing application scenarios. By introducing new instruction formats and optimizing instruction execution efficiency, the performance of the autonomous instruction set can be enhanced. Additionally, a testing and verification system for the instruction set should be established to ensure its stability and reliability.
Build an Instruction Set Ecological System
The success of an autonomous instruction set lies not only in technological innovation but also in the construction of an ecosystem. Companies should actively collaborate with operating system vendors, software developers, universities, and research institutions to jointly build a hardware and software ecological system based on the autonomous instruction set.
In terms of operating systems, collaborate with domestic operating system vendors like Kylin and Tongxin UOS to adapt and optimize the operating system, ensuring it can run stably on the autonomous instruction set architecture. In terms of software applications, encourage software developers to develop application software based on the autonomous instruction set by providing development tools, technical support, and financial subsidies to attract more developers to participate in ecosystem construction. At the same time, establish developer communities and technical exchange platforms to promote experience sharing and technological innovation among developers.
Furthermore, actively explore technologies like binary translation to achieve compatibility with existing mainstream instruction sets. For example, Loongson Technology can use binary translation technology to be compatible with applications on x86/ARM platforms, reducing user migration costs and accelerating the development of the autonomous instruction set ecosystem. By building a complete ecological system, gradually expand the application range and influence of the autonomous instruction set, enhancing its competitiveness in the market.
(2) Microarchitecture: Breaking Through Technical Bottlenecks for Independent Design
Enhance Microarchitecture Design Capabilities
Domestic chip companies should increase R&D investment in microarchitecture design and cultivate and attract a group of experienced microarchitecture design talents. On one hand, collaborate with universities to establish chip design-related majors and courses to train local professionals; on the other hand, actively introduce high-end talents from abroad to inject new technologies and ideas into the companies.
Establish specialized microarchitecture R&D laboratories equipped with advanced R&D equipment and tools. Strengthen cooperation with universities and research institutions to conduct research on key microarchitecture technologies, such as branch prediction, out-of-order execution, and cache hierarchy optimization. For example, in branch prediction technology, research how to improve prediction accuracy, reduce pipeline stalls, and enhance chip performance; in cache hierarchy optimization, explore how to reasonably design cache capacity, structure, and replacement algorithms to reduce memory access latency.
Through continuous technological innovation and practical accumulation, gradually master the core technologies of microarchitecture design, achieve independent design of microarchitecture, and break free from dependence on foreign IP cores. At the same time, establish a knowledge产权保护体系 for microarchitecture design, patenting R&D成果 to enhance the company’s core competitiveness.
Focus on Specific Scenario Optimization
In the microarchitecture design process, it is essential to focus on specific application scenarios for in-depth optimization, considering the unique demand characteristics of the Chinese market. China has the world’s largest new energy vehicle market, the richest AI application scenarios, and the most complete electronic information manufacturing system, providing ample space for optimizing chip microarchitecture.
For example, in response to the demands of the artificial intelligence field, design microarchitectures with strong computing power and high energy efficiency. In the field of autonomous driving, considering the strict requirements for real-time performance and safety, optimize the processing speed and data processing capabilities of the microarchitecture to ensure that the chip can quickly and accurately process sensor data and make correct decisions. By differentiating microarchitecture design, form unique competitive advantages, occupy a place in niche markets, and avoid direct competition with international giants in the general CPU field.
(3) Toolchain: Perfecting the Autonomous and Controllable R&D Support System
Break Through Key Tool Technologies
Domestic companies should collaborate with research institutions and universities to focus on breaking through key technologies in the chip design toolchain. In EDA software, increase R&D investment, form specialized R&D teams, and focus on tackling core technologies in the front-end of chip design, such as logic synthesis and layout routing, as well as back-end physical verification and simulation.
For example, in logic synthesis technology, research how to improve synthesis efficiency and optimize circuit performance, reducing chip area and power consumption; in layout routing technology, explore how to solve high-density routing and signal integrity issues to improve chip reliability. At the same time, strengthen the R&D of tools like compilers and debuggers to build a complete chip design toolchain. Through independent R&D and technological innovation, gradually achieve autonomy and control over EDA software and other tools, reducing dependence on foreign toolchains.
Promote Toolchain Ecosystem Construction
Establish a toolchain developer community to attract more developers to participate in the development and optimization of the toolchain. Through open-source and collaborative approaches, share toolchain technological achievements and accelerate the iterative upgrade of the toolchain. For example, by borrowing the development model of open-source software communities, encourage developers to submit code and propose improvements, creating a positive atmosphere for technological innovation.
Establish close cooperative relationships with chip design and manufacturing companies, continuously optimizing toolchain functions based on actual application needs. For example, adapt and optimize the toolchain for different chip manufacturing processes to improve its practicality and stability. At the same time, actively participate in the formulation of international toolchain standards to enhance the voice of domestic toolchains in the global market. By promoting toolchain ecosystem construction, form an autonomous, controllable, and open innovation toolchain development system to provide solid support for the development of the domestic chip industry.
(4) Supply Chain: Building a Safe and Controllable Supply Network
Strengthen Local Supply Chain Construction
Domestic chip companies should increase support for local supply chains and establish long-term stable cooperative relationships with domestic semiconductor equipment manufacturers and material suppliers. Through investment and technological cooperation, help local companies improve their technological levels and production capabilities.
In semiconductor equipment, support domestic companies in developing key equipment such as photolithography machines and etching machines. For example, Shanghai Micro Electronics has made certain progress in photolithography machine development, and companies should strengthen cooperation with it to accelerate breakthroughs in photolithography technology and its industrial application. In materials, promote technological innovation in core materials such as photoresists and high-purity silicon wafers to achieve domestic substitution of key materials.
At the same time, encourage domestic companies to carry out technological innovation in advanced manufacturing processes, packaging, and testing, gradually improving the local chip industry chain. For example, SMIC has continuously made breakthroughs in advanced process technology R&D, and domestic companies should strengthen cooperation with SMIC to jointly enhance the level of domestic chip manufacturing, reduce dependence on foreign supply chains, and ensure the security and stability of the supply chain.
Diversify Supply Chain Channels
While strengthening local supply chain construction, actively expand diversified supply chain channels. Establish cooperative relationships with companies in semiconductor industry-developed regions such as Europe, Japan, and South Korea, and obtain key equipment, materials, and technologies through technical exchanges and joint procurement.
For example, cooperate with European companies to introduce advanced semiconductor manufacturing technologies and equipment; collaborate with Japanese companies to learn from their experience in semiconductor material R&D and production. Additionally, strengthen cooperation with countries along the “Belt and Road” to jointly build a regional semiconductor industry supply chain. By cooperating with countries along the “Belt and Road,” achieve resource sharing and complementary advantages, enhance the resilience and risk resistance of the supply chain, and reduce risks arising from disruptions in a single supply chain channel.
(5) Industrial Ecology: Cultivating an Innovative Ecosystem for Collaborative Development
Establish an Industrial Ecology Alliance
Domestic chip companies should unite with operating system vendors, software developers, and system manufacturers to establish an industrial ecology alliance. Through the alliance, integrate resources from all parties to achieve technology sharing and collaborative innovation.
Within the framework of the ecological alliance, jointly formulate industry standards to promote compatibility and adaptation of software and hardware products. For example, establish interface standards between chips and operating systems, and application software to ensure that products from different manufacturers can work together and be compatible. At the same time, strengthen cooperation with universities and research institutions to establish a collaborative innovation mechanism between industry, academia, and research. Universities and research institutions have rich research resources and talent advantages, and by cooperating with them, rapidly transform research achievements into actual productivity, enhancing the overall innovation capability of the industrial ecology.
Furthermore, the industrial ecology alliance should actively carry out international cooperation and exchanges, learning from advanced international experiences and technologies to enhance the competitiveness of the domestic chip industry ecology in the international market.
Promote Industrial Mergers and Integrations
Small and medium-sized enterprises are an important part of the chip industry ecology, playing a crucial role in promoting technological innovation and enriching the industrial ecology. Governments and large enterprises should increase support for small and medium-sized enterprises by providing financial support, technical guidance, and market channels to help them solve development challenges.
However, as technology develops to a certain stage, many small and medium-sized enterprises face intense competition in the market, such as in the fields of power management chips and IoT chips, where the market structure is fragmented. It is necessary to guide some enterprises without scale advantages to exit in an orderly manner through industrial guidance funds and listed company merger funds, promoting listed companies to grow stronger.
