On August 16th, the revisions to Japan's "Export Trade Control Order" by the Ministry of Economy, Trade and Industry officially took effect. High-end five-axis machine tools with positioning accuracy of 3 microns or less, along with high-precision rotary tables, optical scales, CNC systems, and even underlying algorithms and manufacturing process parameters, were all shifted from "bulk licensing" to "case-by-case review."
The approval cycle was extended from 15 days to a maximum of 180 days, with a rejection rate exceeding 80% for purchases in sensitive sectors.
Ten years ago, Dong Mingzhu made a statement that nobody took seriously at the time: "One day, we won't be able to buy a single high-end machine tool."
Back then, many who heard this thought "Sister Dong" was being alarmist. Things were being bought just fine; if money could solve the problem, why worry? Today, those words have been validated.
Dong Mingzhu Made the Hard Call, Gree Chewed on a Hard Bone for Over a Decade
In 2013, Gree understood better than anyone what "buying" felt like.
A core component in air conditioner compressors is the impeller, which requires extremely high surface precision and must be processed by five-axis linkage machine tools. At that time, Gree's only options were Germany or Japan.
A single piece of equipment cost tens of millions, and waiting in line for delivery for a year or more was the norm. What's more absurd, sellers installed GPS tracking and remote lockout systems in the machines. If a machine was moved a few centimeters on the workshop floor, the system would lock directly, requiring contact with the supplier's engineers to apply for an unlock and pay another hefty fee.
The other party would casually ask, "You're moving the machine, are you planning to use it elsewhere?"
Something like this would infuriate anyone.
Dong Mingzhu didn't endure it. Around 2013, she made the decision: invest our own money to build machine tools. In 2015, Zhuhai Gree Intelligent Equipment Co., Ltd. was established, and machine tool R&D was listed as a core strategy. A company making air conditioners was going to chew on the hardest bone in industrial machinery—there were many naysayers at the time.
They chewed on it for ten years. All core components were self-developed. Spindle motors, linear motors, CNC systems, five-axis swivel heads, high-precision rotary tables—everything from hardware to algorithms was rewritten line by line.
To conquer the thermal stability of the spindle motor, the team transplanted air conditioner heat dissipation technology, controlling temperature rise to ±1°C and thermal deformation to within 1 micron. Linear motor drives replaced traditional ball screws, achieving acceleration of 1.2G and improving efficiency by over 30%.
In 2017, the first fully self-developed five-axis linkage CNC machine tool, the GA-F500, was introduced. In 2025, a high-speed double five-axis gantry machining center was officially unveiled at the CIMT exhibition, winning the Gold Medal at the 50th International Exhibition of Inventions of Geneva—one of the world's top three invention exhibitions alongside Pittsburgh and Nuremberg, known for its extremely strict judging criteria.
By 2026, Gree's five-axis machine tools accounted for over 75% of export sales. Gree's flagship product is the high-speed double five-axis gantry machining center GA-FMB3020D, specifically targeting the batch processing of large aluminum alloy parts like integrated die-castings, battery trays, and front/rear base plates for the new energy vehicle sector. In this niche segment, its shipment volume ranks among the top three domestically.
Comparing its specifications with similar Japanese models, its paper strength is fully on par:
Gree's model offers rapid traverse speeds of 120m/min, acceleration of 1.1G, spindle speed of 20,000 rpm, repeat positioning accuracy of ±0.012mm, and can complete six-sided machining in a single clamping, with processing efficiency 80% higher than traditional single five-axis machines.
Mazak's gantry five-axis models of similar specifications have slightly higher precision metrics by half an order of magnitude, but do not hold an advantage in aluminum part processing speed or batch production efficiency, while the price is double.
In 2026, Gree's equipment began mass supply to BYD, Tesla, BMW, with exports exceeding 75% of sales. Its cost is only 40-60% of German or Japanese products of equivalent grade, with a delivery cycle of 90 days—compared to a year typically for Japanese models.
Currently, Gree's CNC machine tool product line covers major models including horizontal machining centers, five-axis linkage machining centers, and turning-milling compound machining centers, providing solutions for the 3C, automotive, and mold industries. Objectively speaking, Gree has not yet fully replaced top-tier Japanese equipment.
But Dong Mingzhu's strategic layout has filled a gap in China's industry. Leveraging Gree's manufacturing scenarios, stable cash flow, and continuous iteration, they can step up when being "strangled."
Above 3 Microns, the Puzzle of Domestic Machine Tool Substitution is Largely Complete
Japan's placement of the red line at 3 microns was a precise calculation. In the range from 3 to 50 microns, the localization rate of China's five-axis machine tools is already close to 60%, with the overall localization rate reaching 59.5% in 2025, surpassing imported brands for the first time.
But above 3 microns—the segment requiring higher precision—domestic products still lag. Domestic leaders Kede CNC and Beijing Jingdiao are two representatives, one focusing on "large" and the other on "small," having entered a phase of mass substitution as core players.
Kede CNC's first "ace card" is its high-end CNC system. Its GNC62 is the only domestically developed high-end CNC system with fully self-developed source code. The localization rate of its five-axis complete machines exceeds 90%, and it delivered 1,047 five-axis machine tools in 2025.
Kede CNC does not rely on Siemens or Fanuc, achieving full self-development from underlying algorithms and servo motors to high-precision dual-axis rotary tables.
Beijing Jingdiao specializes in micron-level ultra-precision machining—extremely small dimensions with tolerances of 1-2 microns, which previously only Japanese Makino or Swiss Mikron could handle.
Relying on self-developed high-speed spindle motors of 30,000 to 40,000 rpm and in-machine measurement systems, Beijing Jingdiao achieves continuous machining for hundreds of hours with zero drift. It automatically detects workpiece thermal expansion during cutting and performs micron-level reverse compensation, ensuring "first-part-right."
These three companies each hold their ground, piecing together a complete picture of domestic substitution.
While Kede CNC and Beijing Jingdiao have achieved breakthroughs in laboratories and specific models, reaching mass commercial production, in terms of precision retention over long-term operation, Japan's big three (Fanuc, Yamazaki Mazak, Makino) still hold a generational advantage.
Fanuc's CNC systems still hold a 31.7% market share in the installation volume of China's five-axis system market, with Siemens at 24.8%. Combined, they account for over half. Many domestic machine tool manufacturers build their own complete machines but still have to buy the "brain" from Fanuc.
Gree's strengths lie in processing large aluminum alloy die-castings for new energy vehicles, 3C precision parts, and humanoid robot components. But compared to Japan's big three, there is still a gap of over ten years in the depth of the process database for long-cycle, heavy-duty cutting of extremely hard, difficult-to-machine materials like titanium alloys and high-temperature nickel-based alloys.
Japan's three have been in the machine tool game for nearly a century. Makino's ultra-precision mold machines can achieve mirror-level surface roughness when processing quenched high-hardness mold steel, with precision decay controlled at the micron level even after thousands of hours of continuous operation.
Mazak's heavy-duty five-axis machines, chewing on aerospace-grade hard materials like titanium alloys and high-temperature alloys, set industry benchmarks for tool life and machining stability.
What about Gree? It handles standard molds, 3C thin-wall parts, compressor impellers without issue, but cannot tackle aerospace-grade hard materials like quenched hard molds or high-temperature alloys.
Simply put, Gree has solved the problem of "whether we have it, whether we're being strangled." Japan's decades of accumulated expertise solves the problem of "how stable is it, how long does it last."
Things like precision drift of spindles after 10,000 hours of operation, the adaptability of cutting parameter libraries to different materials, reliability under extreme conditions—there are no shortcuts for these; they can only be built over time.
The good news is that Gree has invested over 2 billion over ten years. Kede CNC is deeply linked to the C919 program, and Beijing Jingdiao's machines have entered aerospace production lines. Domestic has become the only choice.
Taking Kede CNC as an example, its revenue in the first half of 2026 skyrocketed 44.8% year-on-year, with new orders growing 37.5%.
Once the ban took effect, domestic companies accelerated replacing old German and Japanese five-axis machines with domestic equipment, receiving orders "faster than they can handle." Massive amounts of real machining data are rapidly filling the stability gaps of domestic machine tools.
Ten years ago, Dong Mingzhu said, "You can buy equipment, but you can't buy industrial security." Nobody paid attention. A decade later, Japan's control order landed, and everyone realized that the most clear-sighted person had already taken action long ago.
Chinese manufacturing has never lacked people shouting slogans; it lacks people willing to invest tens of billions in places no one sees and chew on hard bones for a decade. Dong Mingzhu's sense of crisis is worth learning for every Chinese manufacturer—don't wait until the knife is at your throat to think about self-development, but prepare for the worst-case scenario when times are at their best.
Japan drew the 3-micron line with precision, aiming to strangle China's machine tool industry. But in doing so, it also lost the 70% market share it had held for over a decade.
Companies like Gree, Kede, and Jingdiao have passed the "can we do it" stage. The next step is to solve capacity ramp-up and long-term stability. There are no shortcuts on this path; they must take it step by step.
Whether for a company or a nation's manufacturing industry, true security never comes from imported purchases. Only by firmly mastering one's own core technology and holding the core things in one's own hands can one sleep soundly.
This article is from the WeChat public account "Hotspot Micro Review" (ID: redianweiping), author: Wang Xinxi
