Nhan Dan Newspaper spoke with Nguyen Anh Tuan, Director of the Viet Nam National Multi-Project Wafer Coordination Center (VNMPW/CC), under the Ministry of Science and Technology; Prof. Dr. Tran Xuan Tu, Director of the Institute of Information Technology at Viet Nam National University, Ha Noi; and Dr. Dang Minh Tuan, Senior Vice President of CMC Technology Group and Director of the CMC Institute of Applied Technology Research.
Selecting products that meet market demand
Q: Over the next five to 10 years, which chip products should Viet Nam prioritise? What criteria should be considered when deciding where to concentrate resources?
Nguyen Anh Tuan: We need to start with the practical needs of businesses using the products and our ability to master them. Viet Nam’s Strategy for the Development of the Semiconductor Industry to 2030, with a vision to 2050, has identified specialised chips as the top priority.
The Ministry of Science and Technology has worked with ministries, sectors, businesses, research institutes, universities, and experts to develop a list of specialised chips prioritised by the State for commissioned research, development, and production. The list comprises around 20 chip categories aimed at strengthening technological capabilities, promoting economic development, and ensuring national security.
Each project will be assessed against five criteria: specific demand and customers; contributions to strategic products and technological self-reliance; the design team’s capabilities and lawful rights to use intellectual property (IP) blocks; manufacturing, packaging, and testing capabilities; and financial and commercialisation plans. Projects must have measurable technical targets, a target price, and a commitment from users to conduct trials.
In my view, priority should be given to microcontrollers; secure Internet of Things (IoT) chips for industrial and measurement equipment; power management and battery management chips for renewable energy and electric vehicles; image-processing and edge AI chips for cameras and smart devices; and security and signal-processing chips for digital infrastructure and telecommunications.
Initially, we should focus on one to three products for which customers are already involved in defining technical requirements and conducting trials. The selected technology must meet performance, reliability, and cost requirements, while the projected production volume must be sufficient to cover development costs.
Prof. Dr. Tran Xuan Tu: The selection of product groups or technological directions must be based on a comprehensive market assessment: What does the market need, and what problems does Viet Nam need to solve? We must also assess domestic capabilities, development potential, and geopolitical factors. From my personal observations, Viet Nam should focus on specialised chips for IoT, edge AI, smart devices, information security, unmanned aerial vehicles, and electronic identification systems.
We should gradually develop our capabilities in design, testing, packaging, and mastering core IP.
Prof. Dr. Tran Xuan Tu
These types of chips are generally manufactured using mid-range process technologies of 32–65 nanometres. They do not require excessively advanced and expensive processes, yet they can address Viet Nam’s needs. We should gradually develop our capabilities in design, testing, packaging, and mastering core IP. Initially, we can work with international foundries to manufacture chips while building up our product development capabilities. This ecosystem will create both inputs and a market for domestic manufacturing facilities in the future. We need to identify which stages we must master ourselves and which should be pursued through international cooperation.
Dr. Dang Minh Tuan: Realistic opportunities lie in chip categories that are within our capabilities, meet domestic demand, and have strategic importance. Viet Nam should not attempt to compete in AI chips for data centres or the most advanced processor technologies, which require billions of US dollars and ecosystems dominated by global corporations.
Three areas could be prioritised: security chips and cryptographic IP, such as secure elements and hardware roots of trust; edge AI chips and neural processing unit IP for smart cameras, IoT devices, transport, and factories; and microcontrollers and systems-on-chip based on the open RISC-V architecture for IoT and digital infrastructure. The open architecture helps avoid dependence on proprietary licences and enables the accumulation of reusable core IP for multiple product lines.
Technology businesses operating AI cameras, cloud infrastructure, and information security systems for customers have a clear understanding of practical needs. They can commission chip development from the specification stage, defining performance, power consumption, target costs, and certification requirements.
Demand can then be validated using prototypes on field-programmable gate arrays (FPGAs), which can be reprogrammed, and by running them in real systems and measuring customer feedback before committing to final design costs and transferring the design for fabrication. This “problem first, chip second” approach helps reduce the risk of producing chips for which there are no buyers.
From research to products
Q: What steps are required to turn a design into a finished product? How should research institutes, universities, and businesses cooperate and share infrastructure to shorten this process?
Dr. Dang Minh Tuan: A chip product typically goes through seven stages: defining requirements and technical specifications; designing the architecture and dividing functions between hardware and software; logic design and verification; creating FPGA prototypes and running software and real-world applications; physical design, finalising the design, and sending it for prototype fabrication through multi-project wafer (MPW) runs, which combine multiple designs on a single semiconductor wafer to share costs; packaging, testing, and reliability assessment; and system integration and certification before mass production.
Software and integration capabilities are advantages of Vietnamese information technology businesses. To be commercially viable, a chip needs to be accompanied by development tools, libraries, and reference solutions.
Dr. Dang Minh Tuan
With their current resources, Vietnamese businesses can master product specifications, architecture, logic design and core IP, verification, FPGA prototyping, embedded software and system integration. Software and integration capabilities are advantages of Vietnamese information technology businesses. To be commercially viable, a chip needs to be accompanied by development tools, libraries, and reference solutions.
For stages requiring substantial investment, businesses should cooperate with international foundries to manufacture wafers, work with electronic design automation (EDA) tool providers, and purchase licences for foundational IP. Packaging and testing can make use of service facilities that have already invested in Viet Nam. Domestically, businesses should work with research institutes and universities to train personnel through real projects and share laboratories, tools, and prototype fabrication runs.
Prof. Dr. Tran Xuan Tu: The biggest gap lies in the ability to turn research results into products that meet technical, reliability, cost, manufacturability, and market-demand requirements. Four gaps need attention: the gap between prototypes and finished products; deficiencies in industrial capabilities and testing infrastructure; the gap between research and business requirements; and shortcomings in technology transfer mechanisms and human resources.
Universities and research institutes should focus on core technologies, design, prototyping, and training. VNMPW/CC should serve as shared infrastructure connecting research with industry. Businesses must participate from the commissioning stage onwards, contributing to co-design, practical requirements, data, testing environments, and customer feedback. International cooperation should aim to transfer knowledge, jointly develop technologies, and strengthen domestic capabilities.
In addition to scientists, Viet Nam needs to develop a workforce of product development engineers who can bridge design, fabrication, testing, and production. The ultimate goal is to establish an ecosystem capable of continuously developing, manufacturing, testing, and improving high-tech products in Viet Nam. The way science and technology projects are designed and evaluated must also change, with continuous funding pathways covering research, prototype fabrication, trials, refinement, and pilot production.
Nguyen Anh Tuan: VNMPW/CC serves as a focal point connecting domestic chip research and design needs with the global semiconductor manufacturing ecosystem. It is establishing a “one-stop” support mechanism to prevent resources from being fragmented. The centre is supported by a group of 21 domestic and international experts.
VNMPW/CC will provide support throughout the process, from research and design to fabrication, packaging, and testing, while also connecting projects with customers. At the design stage, it will facilitate access to EDA tools, IP and process design kits, subject to licensing conditions.
Proposals for pilot fabrication will be assessed for technical and technological soundness and commercial feasibility. Designs using the same process technology will be combined on a single semiconductor wafer under the MPW model and sent to a foundry to share costs.
At present, this stage must still be carried out overseas. The semiconductor plant invested in by Viettel at Hoa Lac Hi-Tech Park is expected to begin pilot production in late 2027. VNMPW/CC will coordinate with the plant to gradually conduct pilot fabrication runs domestically.
After receiving chip samples, VNMPW/CC will support testing, functional and characteristic evaluation, fault analysis, and design refinement. It is investing in laboratories and developing criteria, procedures, and support norms for pilot production, with completion expected in the first half of 2027.
At present, this stage must still be carried out overseas. The semiconductor plant invested in by Viettel at Hoa Lac Hi-Tech Park is expected to begin pilot production in late 2027. VNMPW/CC will coordinate with the plant to gradually conduct pilot fabrication runs domestically.
The laboratory is being funded by the State as infrastructure assets and will operate under a service-provision model. VNMPW/CC has begun negotiations and cooperation arrangements to provide shared IP libraries and EDA tools, standard-cell libraries, process design kits, and the foundry’s design review procedures.
Paving the way for first chip orders
Q: Who can pave the way for the first order for Vietnamese chips? How should risks be shared to ensure quality and competitiveness while making efficient use of public funds?
Dr. Dang Minh Tuan: Securing the “first customer” is the biggest bottleneck. Equipment manufacturers are reluctant to take risks with a chip that has no track record of performance. Technology groups such as CMC can help address this challenge because they have products, infrastructure, and customers.
Three mechanisms could be introduced. First, co-development with a commitment to purchase the output: system businesses work with design teams to establish specifications and commit to purchasing a specified quantity if the chips meet agreed technical targets. Second, testing chips in real-world environments: CMC’s AI cameras, cybersecurity equipment, cloud infrastructure, and data centres could serve as test environments, providing reliability data under actual operating conditions. Third, sharing risks with the State by prioritising products incorporating Vietnamese chips in public procurement, particularly in security and digital government, and establishing a fund or guarantee scheme for first orders.
In return, chipmakers must make commitments equivalent to those expected of professional suppliers. In terms of quality, chips must pass reliability tests in accordance with international standards such as JEDEC. Security chips should be certified against standards such as FIPS 140-3 for cryptographic module security or the Common Criteria for evaluating the security of information technology products.
Regarding delivery schedules, businesses must have a clear product roadmap, commit to long-term supply for five to seven years, and establish alternative supply arrangements. Products should be designed to compete in regional markets from the outset. The first order should serve as a launchpad, not as a basis for long-term subsidisation.
Prof. Dr. Tran Xuan Tu: We need to build both “technical trust” and “commercial trust” in domestically developed chips.
This means establishing independent assessment centres or laboratories to test chips against stringent standards before integration. New chips should be deployed in suitable public-service, industrial, or infrastructure applications within controlled environments. The State should help reduce risks in areas where market incentives are insufficient to attract investment, while businesses remain responsible for their commercial products. Decisions on mass production must be based on product quality and business demand.
Shared centres supporting research, design, fabrication, and testing could be established at major universities, with mechanisms enabling businesses to use their facilities jointly. Products developed could be made available to organisations within the ecosystem at trial prices. Once the products pass the required technical milestones, businesses would invest in and take responsibility for production, integration, warranties, and commercialisation.
Nguyen Anh Tuan: Effective mechanisms must begin with the organisations that use chips. The State and major corporations need to announce the problems to be addressed, technical requirements, testing conditions, and projected demand. The proposal for the list of specialised chips has recommended linking support for research and pilot production to orders under programmes and projects funded by the State budget.
Viet Nam’s semiconductor industry development strategy also identifies a policy of prioritising the use of public funds to procure domestically manufactured electronic equipment.
Viet Nam’s semiconductor industry development strategy also identifies a policy of prioritising the use of public funds to procure domestically manufactured electronic equipment.
Nguyen Anh Tuan
I propose a two-stage approach. During the technology development stage, the State should support shared infrastructure and part of the costs of design, prototype fabrication, and testing. Businesses should contribute personnel, technology, and capital, while customers participate in defining specifications, providing testing equipment and co-funding the work. Funding should be disbursed according to milestones. Technical risks must be anticipated and independently assessed. If a testing round fails, the causes should be identified so that the project can be adjusted or discontinued.
During the commercialisation stage, customers could commit to purchasing the first batch, provided the chips pass testing and meet requirements for quality, cost and supply capacity. Subsequent orders should depend on actual performance in use.
To ensure efficient use of public funds, project requirements must be made public, selection must be competitive, assessments must be independent, and support levels and counterpart funding requirements must be specified to avoid duplicated funding. Contracts must clearly define intellectual property ownership, rights to use the results, and the responsibilities of all parties. VNMPW/CC will provide technical assessments, while decisions on commissioning and procurement remain the responsibility of the competent authorities and businesses in accordance with the law.
Measuring capabilities by 2030
Q: Looking ahead to 2030, what achievements would demonstrate Viet Nam’s ability to master semiconductor technology and bring Vietnamese semiconductor products to market?
Nguyen Anh Tuan: I would look for at least one specialised chip product line designed and mastered by a Vietnamese business, integrated into commercial equipment, and generating repeat orders. The business must master the core design, hold lawful rights to use the IP, and be able to fix defects independently and develop subsequent versions. Evidence of market success would include contracts, delivered products, revenue, repeat orders, and the value retained domestically after costs.
Prof. Dr. Tran Xuan Tu: Progress should be measured by products that have passed through the entire chain of design, fabrication, testing, integration, and pilot production, and have customers using them. How many chips designed and mastered in Viet Nam have been fabricated, independently tested, integrated into commercial products, and generated revenue? One successful chip is not enough to demonstrate industrial capability. Viet Nam needs teams and infrastructure capable of continuously delivering second and third products with improving development times, lower costs, and higher success rates.
Dr. Dang Minh Tuan: I would also look for at least one chip product line designed and owned in terms of intellectual property by a Vietnamese business. One example could be an RISC-V-based security chip that obtains international certification and achieves commercial shipments of at least one million units, including export orders. Production volumes, certification, and revenue would provide evidence that the product sells, earns market trust, and retains knowledge and intellectual property in Viet Nam.
Reporter: Thank you very much!
