韩国5000亿韩元SiC功率半导体研发项目停滞，大企业犹豫不决
韩国政府推动的5000亿韩元SiC功率半导体研发项目进展停滞，因主要企业犹豫不决。该项目旨在培育下一代功率半导体产业，但缺乏企业参与。
原文深度解读
本文报道韩国政府推动的5000亿韩元SiC（碳化硅）功率半导体研发项目因主要企业犹豫不决而进展停滞。文章涉及韩国政府、韩国半导体产业协会、贸易产业能源部及主要企业，时间点为2024年6月至8月。关键数字包括5000亿韩元预算、90%-95%的功率半导体进口依赖、约10%的SiC/GaN技术自给率、2030年20%的自给率目标。该项目旨在培育下一代功率半导体产业，与电子元器件供应链中的功率半导体国产化、供应链稳定密切相关。
- 政府于6月12日紧急经济长官会议上宣布将在6月内确定“新一代功率半导体商业化技术路线图”，并开始规划与需求企业联动的大型研发项目。
- 截至8月初，路线图尚未正式公布，需求企业联动研发停留在任务规划和预算协商阶段，未进入具体任务公告或承包方选定。
- 国内功率半导体需求约90%-95%依赖进口，SiC和GaN等下一代宽禁带半导体技术自给率约10%。
- 政府路线图目标是到2030年将技术自给率提升至约20%，核心战略包括覆盖材料、器件、模块、系统的全周期商业化技术、需求企业联动的大型研发、利用公共晶圆厂和示范基础设施。
- 行业人士认为，由于中国低价化合物半导体攻势加剧、功率半导体业务盈利能力相对较低，难以吸引大企业参与。
- 产业通商资源部官员表示正在与企业持续协商，具体时间难以确定，但详细工作完成后将尽快公布。
供应链影响
- 韩国功率半导体国产化进程可能延迟，导致对进口的依赖持续，供应链稳定性可能取决于政府与企业协商的进展。
- 若大企业不参与，SiC等下一代功率半导体的本地生产能力可能不足，影响电动汽车、AI数据中心等关键行业的供应链安全。
- 政府若加强“控制塔”角色，可能加速项目推进，但具体效果取决于政策执行力度和企业响应。
- 中国低价化合物半导体的竞争可能继续压制韩国本土功率半导体业务的盈利能力，进而影响投资决策。
- 技术自给率提升目标（2030年20%）的实现可能取决于能否吸引足够的需求侧企业参与和公共基础设施的有效利用。
系统解读，仅作为判断线索，不构成备货、出货、涨价或投资建议。
商业化路线图公布推迟
企业因盈利能力低而犹豫参与
“政府需要发挥控制塔作用”
一位参观者在电气、电子及材料元件展上查看Trino的高压功率半导体晶圆。[ETnews DB]
政府为培育下一代功率半导体碳化硅（SiC）产业而推进的5000亿韩元研发项目，因未能确保拉动需求的核心企业而难以取得进展。由于大企业以盈利能力和商业可行性低为由表现出消极态度，政府以与需求侧企业联动为前提的研发项目从规划阶段就遭遇挫折。
政府于6月12日在紧急经济长官会议上宣布，将在6月内确定“新一代功率半导体商业化技术路线图”，并开始规划与需求侧企业联动的大型研发。计划通过让包括大企业在内的需求侧企业作为核心企业参与，细化具体任务和项目结构。
然而，截至8月初，尚未发布确认路线图的正式公告。与需求侧企业联动的研发也未能推进到具体任务公告或承包方选定阶段，仍停留在任务规划和预算协商中。
业内人士认为，尽管政府和韩国半导体产业协会正在说服大企业参与，但未能确定实质性的核心企业成为项目推进的瓶颈。分析显示，由于中国低价化合物半导体的攻势加剧，功率半导体业务的盈利能力相对较低，难以吸引大企业参与。
国内功率半导体需求约90%至95%依赖进口，SiC和氮化镓（GaN）器件——下一代宽禁带半导体——的技术自给率仍维持在约10%。为了增强电动汽车、人工智能（AI）数据中心、可再生能源、电网和国防等关键产业的供应链稳定性，国产化是必要的，但能够处理大规模需求的企业参与并未跟上。
政府功率半导体培育路线图的目标是到2030年将技术自给率提升至约20%。确保覆盖材料、器件、模块和系统全生命周期的商业化技术，以及涉及需求侧企业的大型研发，并利用公共晶圆厂和示范基础设施，是战略的核心。
一位行业官员表示：“个别中小型企业正在单独联系大企业以说服他们参与，但很难产生实质性成果，”并补充说：“除非政府直接介入，否则这与‘撞墙式’销售无异。”
有反应认为，在SiC制造现场以及设计方面，这是一个难以诱导大企业参与的业务。虽然同意供应链多元化和国产化的必要性，但利润率微薄导致业务吸引力低。在先进工艺或内存业务等投资回报率高的需求旺盛的情况下，很难将人员或内部资源分配给功率半导体。对于所需数量，从外国公司进口和使用产品仍然更高效的判断也在起作用。
有人指出，为了加快项目进度，政府应发挥主导作用，扮演“控制塔”角色，而不是将联盟组建完全留给私营部门。分析显示，各实体在预算削减、盈利能力和政策成果方面坚持自身立场是瓶颈现象的根源。事实上，在SiC基础设施完善的国家，也采用了让行政单位相互竞争以产生项目成果的政策。
主管部委产业通商资源部的一位官员表示：“我们目前正处于与企业持续协商的阶段，”并补充说：“很难确定具体时间，但详细工作完成后将尽快公布。”
Commercialization Roadmap Announcement Delayed
Companies Hesitate to Participate Due to Low Profitability
“Government Needs to Act as a Control Tower”
A visitor at the Electrical, Electronic & Material Components Fair examines Trino's high-voltage power semiconductor wafer. [ETnews DB]
A KRW500 billion research and development (R&D) project pursued by the government to foster the silicon carbide (SiC) industry, a next-generation power semiconductor, is failing to gain momentum because it has failed to secure anchor companies to drive demand. As major corporations show a passive attitude citing low profitability and commercial viability, the government's R&D project, which is premised on linking with demand-side companies, is suffering setbacks from the planning stage.
The government announced at the Emergency Economic Headquarters Meeting on June 12 that it would finalize the “Next-Generation Power Semiconductor Commercialization Technology Roadmap” within June and begin planning large-scale R&D linked with demand-side companies. The plan was to flesh out detailed tasks and the project structure by having demand-side companies, including major corporations, participate as anchor companies.
However, as of early August, no official announcement confirming the roadmap has been made. The demand-side company-linked R&D has also failed to move forward to the stage of specific task announcements or contractor selection, remaining stuck in task planning and budget consultations.
Industry insiders view that although the government and the Korea Semiconductor Industry Association are persuading major corporations to participate, the failure to finalize substantial anchor companies is acting as a bottleneck in advancing the project. Analysis shows that it is difficult to draw the participation of major corporations because the offensive of low-priced compound semiconductors from China is intensifying and the profitability of the power semiconductor business is relatively low.
Domestic demand for power semiconductors relies on imports for about 90% to 95%, and the technology self-reliance rate for SiC and gallium nitride (GaN) devices—next-generation wide-bandgap semiconductors—remains at around 10%. Localization is necessary to enhance the supply chain stability of key industries such as electric vehicles, artificial intelligence (AI) data centers, renewable energy, power grids, and defense, but the participation of companies capable of handling large-scale demand is not supporting the effort.
The goal of the government's power semiconductor nurture roadmap is to raise the technology self-reliance rate to around 20% by 2030. Securing commercialization technology covering the entire lifecycle of materials, devices, modules, and systems, along with large-scale R&D involving demand-side companies, and utilizing public fabs and demonstration infrastructure form the core of the strategy.
An industry official said, “Individual mid-sized and small and medium-sized enterprises are contacting major corporations individually to convince them to participate, but it is difficult to produce tangible results,” adding, “Unless the government steps in directly, it is no different from 'banging one's head against the wall' sales.”
Reactions are emerging that it is a business where it is difficult to induce the participation of major corporations in SiC manufacturing sites as well as design. While agreeing on the need for supply chain diversification and localization, it means the business attractiveness is low due to thin profit margins. Under circumstances where demand for advanced processes or memory businesses with high return on investment is large, it is difficult to allocate personnel or internal resources to power semiconductors. The judgment that it is still more efficient to import and use products from foreign companies for required volumes is also playing a role.
Points are being raised that for the project to speed up, the government should take the lead and play a “control tower” role rather than leaving consortium formation entirely to the private sector. Analysis shows that individual entities insisting on their own positions regarding budget cuts, profitability, and policy results are the cause of the bottleneck phenomenon. In fact, in countries where SiC infrastructure is well established, policies that pit administrative units against each other in competition to produce project results are also being utilized.
An official from the Ministry of Trade, Industry and Energy, the ministry in charge, said, “We are currently in a stage of continuous consultation with companies,” adding, “It is difficult to specify an exact timing, but we plan to announce it as soon as the detailed work is completed.”