Author: Peng Cheng
If you have been closely monitoring semiconductor industry recruitment websites and business registration information over the past two years, you have likely repeatedly encountered unfamiliar names: Aiguang Yintu, Shenwei Zhichuang, Hengsheng Tianji, Yuanzi Qizhi, Langdao Suiyan, Jingzhi Xin...
Almost all of them were established after 2024, with registered capital ranging from tens of millions to CNY 5.5 billion. The shareholders are not ordinary venture capitalists, but rather Zhangjiang Group, Pudong State-owned Assets, ChangXin Industrial Investment, SMIC Capital, Vital Materials, etc. Most of these companies have yet to launch products, yet they have already begun large-scale hiring or publicly unveiled prototypes with a 90% domestic localization rate.
Semiconductor equipment is a typical long-cycle, high-investment, low-tolerance-for-error industry. It often takes more than five years for a device to move from prototype to mass production introduction. Why have so many teams dared to enter simultaneously in just these two years? This is not the usual story of "startup teams securing funding," but rather an organized and rhythmic formation of forces.
01New Faces in the Semiconductor Equipment Sector
Combining business registration information with public reports reveals that at least ten major new semiconductor equipment entities have entered the public eye since the second half of 2024. An incomplete list follows:
It is evident that the companies on this list share three common characteristics. First, most are led or deeply involved by state-owned capital or large enterprises, while purely market-driven VC-backed companies are actually in the minority. Second, their directions are remarkably consistent, targeting lithography, metrology and inspection, light sources, advanced etching, etc., which are almost entirely segments with low domestic localization rates. Third, capital precedes product; Shenwei Zhichuang registered CNY 5.566 billion in capital without having built a single device, which is extremely rare in the equipment industry.
02Validation Opportunities Are the Bottleneck for New Semiconductor Equipment Vendors
In the semiconductor equipment industry, moving from R&D to mass production shipment has never been hardest because of building the prototype. Chambers, optics, motion stages, and control software can be gradually developed if there is money and talent. The real difficulty lies in how to obtain validation opportunities from wafer fabs, because no wafer fab is willing to allow a device that has not run in mass production onto its production line.
In the early years, domestic high-end manufacturing capacity was highly concentrated, with only a few customers purchasing semiconductor equipment. These major players primarily focus on yield, capacity, and utilization rate metrics. Introducing a new machine onto the line poses risks such as particle exceedance, process drift, and scrapping of entire batches of wafers, which process and equipment managers find difficult to bear. Mature imported equipment is stable, closed-loop, and maintainable, giving no one reason to risk switching devices. Thus, a vicious cycle formed: without machine access, no process data could be generated; without process data, customers were afraid to use it; without orders, there was no revenue to support the next iteration. This cycle kept most teams out and left segments like lithography light sources and electron beam metrology with only a few long-term players, making it difficult for new companies to enter. Therefore, the focus of this round of change is not on the equipment companies themselves, but on how to secure validation resources.
03Formation of Validation Systems: Layered and Progressive Open Platforms
In recent years, a multi-layered validation system has gradually emerged in China. Several mechanisms have formed simultaneously, each accommodating different R&D stages of prototypes.
Open shared pilot lines within public process platforms are the most critical starting point for startup equipment vendors. The primary goal of mass-production wafer fab lines is stable output, yield, and delivery; whereas the primary goal of pilot platforms is process development, equipment-on-machine validation, material adaptation, and small-batch tape-outs, allowing repeated tuning, process changes, and compatibility with multiple routes, offering far greater error tolerance than mass-production fabs.
In earlier years, the supply of public pilot validation in China was scarce. Startup equipment vendors queuing directly at mature mass-production fabs had almost no chance; this was the "validation deadlock" for domestic equipment startups in the past. In recent years, various regions have concentrated on launching and opening a batch of public pilot validation platforms around third-generation semiconductors, silicon photonics, MEMS, and specialty processes, forming a "validation buffer layer" independent of mass-production fabs.
On one hand, this provides startups with low-cost, implementable machine-access scenarios to first produce validation data for external display; on the other hand, it reduces risk for wafer fabs, as the platform has already completed a round of screening, meaning fabs do not need to bear trial-and-error costs from scratch. Taking the 8-inch line at Pinghu Laboratory as an example, it opens up to SiC/GaN wide-bandgap semiconductors, specifically undertaking on-machine validation for new domestic equipment and materials, having already validated hundreds of domestic pieces of equipment.
In addition to the aforementioned public platforms, newly built wafer fabs reserve domestication validation stations during the project initiation phase. In recent years, a batch of newly built memory, specialty process, power, and advanced packaging lines have reserved isolated validation positions that do not carry mass-production tasks when planning cleanrooms, tool layouts, and gas/power distribution designs. The assessment target for these stations is not yield, but the progress of domestication validation. Take CR Micro as an example; focusing on the power device specialty process track, it prioritizes full-chain introduction of domestic equipment, leveraging the larger error tolerance of power devices compared to advanced logic chips to build specialty process lines dominated by domestic equipment. Targeting IGBT and MOSFET process requirements, it conducts targeted bidding for domestic etching, ion implantation, thin film, and cleaning equipment vendors to carry out joint process development; it includes domestic equipment in the initial equipment list during the new line planning stage, rather than replacing them after construction, reducing later retrofitting costs; it collaborates with equipment companies to optimize equipment modifications for special power device processes; and it accumulates experience using domestic equipment into internal standards, replicating mature procurement and introduction solutions across multiple subsequent lines.
Risk guarantee mechanisms help enterprises reduce validation failure costs. In the past, resistance to new machines on production lines stemmed partly from liability concerns regarding scrapped wafers. Over the past two years, various regions have introduced a series of guarantee mechanisms for semiconductor equipment validation to help enterprises lower validation losses. Taking Xiamen as an example, to solve the marketing difficulties of domestic chips and equipment being "too risky to use" or "unwilling to be used," new policies established a risk-sharing mechanism. It encourages manufacturing and OSAT lines to introduce approved first-set equipment and first-batch materials. For those that pass validation and achieve mass production, substantial subsidies are provided equivalent to 50% of the actual costs incurred by the introducer, with a maximum annual subsidy of CNY 5 million per enterprise, directly lowering the cost for downstream manufacturers to test the domestic supply chain. Additionally, the Ministry of Industry and Information Technology (MIIT), the Ministry of Finance, and the National Financial Regulatory Administration jointly launched an insurance compensation mechanism for major technical equipment and new materials in their first set/batch. For qualifying projects, central finance can provide insurance compensation of up to 80% of the actual premiums paid, breaking the cycle of "no large-scale validation for domestic equipment → users dare not purchase → industrialization hindered."
04More Buyers, Changed Coordination Methods
The surge of enterprises entering semiconductor equipment is driven not only by upgraded validation mechanisms but also by changes on the demand side: the number of downstream entities capable of buying equipment has increased, and they are no longer the same group of people. This allows the market to support more equipment companies.
First, AI is the most direct driver of this wave of demand. The expansion of memory and advanced logic capacity driven by HBM and AI servers has exploded, leading to significant growth in capital expenditure by domestic wafer fabs. Equipment is a prerequisite for capacity expansion; before wafer fabs are even built, orders arrive at equipment factories. Rising demand automatically brings spillover of capital and talent: profitability in the equipment sector surges, with about 70% of the CNY 344 billion capital of Big Fund Phase III directed toward equipment and materials, pushing engineer salaries up by a tier overall. When a track simultaneously possesses certain demand, policy support, and profit effects, establishing an equipment company shifts from a gamble to a natural progression. More critically, export controls combined with 12–18 month lead times for overseas equipment mean new wafer fabs cannot wait for imported tools, forcing them to give validation and trial-and-error opportunities to new domestic equipment.
Second, numerous brand-new entities have appeared among domestic new wafer projects landing in these two years, including new-brand memory IDMs, locally built specialty process plants, power and compound semiconductor lines, and advanced packaging lines. They are independent of old giants in terms of equity, teams, and process routes, naturally motivated to accept new domestic suppliers: project approvals, matching funds, and syndicate agreements often come with domestication quotas; new fabs need to compete with mature giants for orders, yield, and cost, making it difficult to differentiate by simply copying the imported equipment lists of old fabs; they are also unwilling to be locked into delivery, maintenance, and spare parts by a single supplier.
Purchasing power has dispersed from a few familiar faces to a batch of new players, effectively expanding the capacity of the equipment customer pool.
Finally, there is synergy within the same capital lineage. A common practice now is that the same local integrated circuit fund or state-owned asset platform invests in building new wafer fabs while simultaneously diversifying investments in supporting equipment and component companies. On paper, they appear independent with no mandatory related-party transactions; however, at the post-investment management level, they can coordinate validation slots and schedules, adjust validation costs between project pools, organize closed-door process alignment meetings, and prioritize invested equipment for local domestic alternative lists. The Shenzhen state-owned asset system places equipment platforms and multiple wafer fabs (covering advanced nodes, mature nodes, and memory) on the same capital platform, completing validation within a closed loop inside the system; Zhangjiang Group wholly owns Shenwei Zhichuang while holding equity in Shanghai Micro Electronics and its spun-off platforms through subsidiaries, binding equipment and industrial platforms on the same line.
Shanghai even explicitly stipulates that production lines and pilot lines providing validation services for domestic first-set equipment can receive subsidies, up to CNY 1 million per unit; meanwhile, new integrated circuit production line projects seeking government financial support must ensure that the proportion of autonomous equipment and material purchases is no less than 30% of the supported funds. Beyond local IC funds and state-owned assets, some downstream customers are also investing in upstream equipment, locking in capacity in advance while providing validation opportunities for equipment vendors. However, related relationships only make it easier for prototypes to enter the field; hard metrics like particle control, long-term MTBF, and spare parts response time will not relax in the slightest.
05Talent, Capital, and a Narrower Path
Beyond validation mechanisms and exploding market demand, the spillover of talent and capital has also driven the emergence of this batch of semiconductor equipment vendors.
The conditions on the talent side only truly matured in the last two years. From 2018 to 2022, domestic equipment underwent scaled expansion. Companies like Advanced Micro-Fabrication Equipment, Naura Technology, ACM Research, Piotech, and Kesiton trained a whole batch of professionals who understood complete machines, processes, vacuum RF, precision motion, and optics through successive projects. Combined with the contraction of some overseas giants' domestic teams and adjustments in project pacing, key personnel with complete project resumes began flowing outward. In the early days, assembling a team was the biggest headache for startups; since 2024, this has shifted from scarcity to feasibility.
On the capital side, the logic has also changed. In earlier years, equipment investment was mostly policy-driven with distant stories; no one dared to touch a track with no revenue for seven or eight years. This round is different: downstream has verifiable expansion expectations, and capital can internalize the validation link through industrial portfolios, thus becoming willing to diversify bets on niche segments rather than betting solely on a few top players. Many funds even allow equipment companies to have no revenue for several consecutive years.
Product route choices for equipment companies also lean towards niche markets. The vast majority of new companies avoid touching relatively mature domesticated markets like etching and thin film deposition, instead cutting into single points: electron beam metrology, light source accessories, vacuum subsystems, special cleaning, bonding, dedicated process inspection, etc. This avoids head-to-head competition with mature leaders and naturally matches the new validation ecosystem; single-point equipment has simple process interfaces and shorter R&D cycles, making it easier to get feedback first from public platforms or specialty process lines.
Regional division of labor has also differentiated synchronously. Shanghai remains the densest location, with the most complete trio of talent, supply chain, and funds, tackling the most difficult segments; Wuxi, Hefei, Suzhou, and Wuhan focus more on supporting local wafer and OSAT lines, securing validation opportunities from local production lines first.
06Conclusion
It is evident that domestic equipment is transitioning from "single-point breakthroughs by a few leaders" to the second stage of "cluster complementation across the entire supply chain." The first wave (2018—2022) solved the problem of going from zero to one for mature main equipment; the second wave (since 2024) attacks the niche bottlenecks and key subsystems that independent teams found difficult to touch in the past. What pushed the formation of the second wave was the simultaneous establishment of four things: layered validation channels, diversified buyer entities, genealogical industrial capital, and normalized supply of talent and long-cycle capital. Only when these four conditions overlap does it explain why this is the specific time window, neither earlier nor later.
However, this boom comes with inherent fragility. In the same niche segment, three or four teams often emerge simultaneously with highly similar product definitions, where the ultimate battle is who secures stable process data first. An increase in validation channels does not equate to a synchronous rise in pass rates; performing well on public lines does not guarantee survival in mass-production environments regarding long-term stability, particle control, O&M costs, and spare parts delivery, with many projects potentially stalling at the pilot stage. Teams overly reliant on related production lines to get their first ticket easily fall into the customization trap: adapting only to a single process leads to obstacles everywhere when selling externally, as customized prototypes are not equal to general-purpose mass-production equipment. More realistically, new wafer fabs themselves face risks of delay or suspension; once the production line rhythm changes, sample validation tied to them is directly interrupted.