Many people believe that the open-source RISC (Reduced Instruction Set Computing) architecture will determine the future of chip design and the development of new technologies like next-generation artificial intelligence. RISC-V is an open standard developed through international cooperation, and participating in such international standards is thought to help companies maintain greater control over their intellectual property and promote innovation in both the public and private sectors. However, some U.S. policymakers are concerned that the RISC-V architecture standard may threaten U.S. national security and competitive advantage. This article explores the potential impact of the RISC-V architecture on the global chip landscape through seven questions related to RISC-V.
Question 1: What is RISC-V?
The Instruction Set Architecture (ISA) defines how software controls processor hardware. It guides the chip in processing data and performing memory operations, allowing chip designers to create their own chips based on the ISA. Currently, specialized AI chip design companies like Nvidia typically design their own custom ISAs, while general-purpose computing chips (Central Processing Units, or CPUs) tend to adopt existing ISAs because it is more cost-effective, provides better software compatibility, and ensures higher reliability.
There are currently three main semiconductor ISAs: Intel/AMD’s x86, ARM architecture, and RISC-V architecture. ARM architecture and Intel/AMD’s ISA are proprietary, and other companies can use them by obtaining intellectual property (IP) licenses. In contrast, RISC-V is an open standard ISA that allows chip design teams to implement the standard in various situations without paying expensive licensing fees.
RISC-V was originally developed at the University of California, Berkeley, and is now managed by the nonprofit standards organization RISC-V International based in Switzerland. RISC-V International has attracted over 4,000 members from 70 countries, including many U.S. companies, universities, community organizations, and individuals who collaborate to establish the rules for the RISC-V ISA.
It is worth noting that the RISC-V standardization does not require U.S. companies to share confidential information with others. Sensitive IP from companies is used solely for the implementation of RISC-V and is not shared through the RISC-V platform itself. RISC-V does not contain sensitive IP, and its collaborative development does not require companies to disclose such IP; the main feature of RISC-V is that its use is open and royalty-free.
Question 2: What are the advantages of RISC-V?
As an open standard ISA, RISC-V can be implemented for free and is easily accessible. RISC-V allows companies of various sizes to build their implementations based on the ISA, enabling them to better control their technology and leverage the software compatibility of the entire RISC-V ecosystem. Many companies that currently obtain ISA licenses from Intel/AMD or ARM architecture are helping to develop RISC-V to provide an alternative option for their semiconductor architectures. The customizable, low-cost ISA solutions are particularly attractive to small companies.
Some U.S. companies have already recognized the importance of RISC-V and are rapidly investing in software implementations based on RISC-V. In fact, in May 2023, industry leaders, including Google and Nvidia, launched the RISC-V Software Ecosystem (RISE) project aimed at accelerating the application of RISC-V software in consumer electronics, data centers, and automotive products. Even Intel has shown strong interest in advancing RISC-V to support its future chip designs.
Question 3: Why is RISC-V controversial?
In 2022, the U.S. imposed export controls on certain chips and chip manufacturing equipment to specific countries. However, these restrictions did not include chip architectures. In 2023, some U.S. congressional members expressed concerns about the implications of companies from specific countries participating in the development of RISC-V for national security. Some policymakers worry that companies from certain countries might use the RISC-V architecture to circumvent U.S. export controls and could potentially use RISC-V in high-performance chips in the future, even though these chips do not primarily use the RISC-V architecture at present.
Question 4: What does RISC-V mean for the leadership of U.S. chip design?
Currently, chips designed in the U.S. lead the global market in performance, energy efficiency, complexity, and market share. RISC-V offers an alternative platform that allows other countries, companies, and individuals to design chip architectures without the intellectual property and cost constraints of Intel/AMD and ARM architectures.
In fact, RISC-V International’s move overseas was to avoid potentially violating U.S. sanctions against companies from specific countries (i.e., RISC-V members). This means that the U.S. cannot restrict the release of the RISC-V standard or decide which countries can participate in this process. The only direct but not necessarily effective way is to limit U.S. companies’ participation in the development of the RISC-V standard.
Question 5: What would happen if U.S. companies were restricted from participating in RISC-V?
Restricting U.S. companies from participating in RISC-V could have counterproductive effects. In this case, RISC-V International might continue to develop without relying on U.S. participation. U.S. companies would be forced to develop technologies suitable for different international integrated circuit standards, license expensive proprietary standards, or adopt less efficient technologies. Ultimately, U.S. companies might be excluded from the development of RISC-V standards, relinquishing their seats at the negotiating table and losing influence over their development trajectory.
Importantly and challengingly, RISC-V has already been integrated into public and private sector projects within the entire U.S. innovation ecosystem, from NASA’s Jet Propulsion Laboratory’s spaceflight computing to DARPA’s hardware architecture security. One significant advantage of this shared collaborative ‘language’ is that researchers can leverage it, universities can teach it, and both small startups and large companies can adopt it; for example, they can build new chips to support the next generation of artificial intelligence. Therefore, the U.S. has ample reason to continue closely participating in the development of RISC-V standards.
Question 6: How do open standards like RISC-V work?
Global standard organizations typically decide how technologies are developed. Open standards provide a pathway for industry participants to collaboratively develop technological solutions. Technical standards act like a language, setting parameters for communication and interoperability. Shared and collaborative standards help promote innovation and limit abuse.
Historically, U.S. leadership in global standardization has fostered growth and innovation in technology and its supply chains, providing corresponding advantages to U.S. companies. Now, specific countries have also recognized the value of actively participating in standardization to develop their long-term competitive advantages.
Question 7: How does RISC-V challenge U.S. economic security?
RISC-V enables chip designers (whether allies or adversaries) to reduce their dependence on the ISAs of Intel/AMD and ARM architectures. This presents both opportunities and challenges for the U.S. chip design ecosystem, creating a crossroads for U.S. economic security.
On one hand, U.S. policymakers may restrict U.S. companies’ participation in RISC-V and support the domestic design industry through subsidies or tax breaks. The aim is to weaken RISC-V’s overall competitiveness compared to companies like Intel.
On the other hand, U.S. policymakers may further push for U.S. participation in RISC-V, embracing the international standardization process and investing in national standards education to consolidate U.S. leadership and influence in the global chip design arena.
Currently, standards are being developed globally for many emerging technologies, including artificial intelligence. For example, discussions around RISC-V reflect debates regarding open-source foundational models, where the choice is between regulatory control or competitive control. Like RISC-V, open-source foundation models will continue to surge. The U.S. may choose to restrict their development or opt to support them.