A Chinese team turns a five-axis CNC machine tool into a desktop product: overseas crowdfunding explodes, but export controls limit it

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

English text is being prepared.

You can think of it as an automated miniature metalworking shop. If you want to make robot joints, watch cases, or small impellers, you just clamp a piece of metal inside, import the model you've drawn, close the hatch, and the machine will slowly cut it into the required shape from different angles.

Take the WM-100 as an example. Before machining begins, the machine first scans the position of the material; during machining, it can also automatically switch among six tools: first use a large tool to cut out the outline, then switch to a small tool to refine the details, and finally switch to a drill bit to make holes. What used to require workers to stop the machine again and again, change tools, and recalibrate can now be completed continuously by the machine.

So the people most interested in this type of machine are mainly hardware startup teams, robotics companies, university laboratories, makerspaces, and small machining workshops. They usually don't need hundreds or thousands of identical parts; their needs are typically very fragmented: a robotics team wants to try out a new joint, a drone company needs to modify a connector, a lab temporarily needs a set of fixtures, and a design studio wants to make a few irregular enclosures.

If these jobs are handed to a large factory, the quantity is too small, and just requesting quotes, scheduling, and communicating takes a long time. When the sample comes back and they find that one hole position is off, the whole process has to be repeated. With a desktop five-axis machine, engineers can revise the drawing in the morning and try cutting in the afternoon, then adjust the dimensions if it doesn't fit.

Many people also buy it purely for their own hobbies. In the United States, there is a pair of mechanical engineer brothers who often used the school's machine tools to make things while studying. After graduating, they no longer had the equipment, so the two of them jointly bought a desktop five-axis machine tool. They once spent four or five hours making a titanium alloy ring, and later used it to machine watch cases, wrenches, and jewelry, essentially moving the school laboratory back home.

In China, a programmer named "Zishu" likes cave diving. When he couldn't buy suitable device accessories on the market, he designed them himself. Previously, he had to send the drawings to a factory and then rush to test after the parts were made; after buying a desktop CNC, he could complete the machining at home. Later, he also sold the finished accessories to other diving enthusiasts, selling more than 20 pieces in total.

Pretty cool, right! But for overseas users who want to buy such a five-axis simultaneous-motion desktop machine tool from China, it is not so easy.

The WM-100 project team announced that this round of crowdfunding is open only to companies, factories, universities, and research institutions, and requires buyers to provide institutional registration information, an institutional introduction, and a commitment letter on final use. Ordinary users are temporarily not eligible to purchase.

The reason is that the WM-100's technical specifications fall within the scope listed in the relevant Nuclear Suppliers Group lists, and exports must apply for a license in accordance with China's dual-use item export control regulations. Export controls look at what the machine can machine and how high its precision can be, not how much space it occupies. So even though its size and price are already close to consumer-grade products, it cannot be sold overseas directly like ordinary appliances.

Foreigners who cannot place orders are anxious and only hope that individual user purchases will be opened soon; they even think the $6,000 price is "dirt cheap."

So why would a desktop device that can only machine palm-sized parts trigger nuclear-related export controls?

In July 1987, outside the U.S. Capitol, several members of Congress swung sledgehammers and publicly smashed a Toshiba radio.

The radio certainly had done nothing wrong. What really angered them was that a machinery company under Japan's Toshiba had secretly sold several CNC machine tools to the Soviet Union.

Just a few machine tools—was it really worth such an uproar? It really was.

At the time, the Cold War was in full swing, and the United States and the Soviet Union were engaged in fierce underwater competition. The U.S. Navy had an important advantage: some Soviet submarines were noisy, and the U.S. military could use sonar to capture the sound and track their positions. The Soviet Union naturally wanted its submarines to be quieter, and the propeller was one of the areas that needed improvement.

When a submarine is underway, the propeller rotates at high speed, and a large number of tiny bubbles may form around it. When these bubbles burst, they make noise, leaving sonar with clues for tracking. To reduce noise, it is necessary to improve the design and manufacturing precision of the propeller.

The trouble is that submarine propellers are not ordinary electric fan blades. Their diameter can reach several meters, the blades are thick and large, and the surface has complex curves and twists. To accurately mill such shapes out of metal materials requires extremely capable machining equipment.

High-performance multi-axis CNC machine tools are the key among them. For this reason, during the Cold War, the Coordinating Committee for Multilateral Export Controls (COCOM), established by Western countries, placed such advanced equipment on the embargo list against the Soviet Union.

But in 1981, the Soviets still bypassed this gate. Through a trading company, they found Toshiba Machine, under Japan's Toshiba, and purchased four nine-axis CNC milling machines.

Nine axes—several more independently controllable motion axes than the five axes we discussed earlier. This batch of machines was also astonishing in size, standing two stories high and weighing more than 200 tons, specifically used for machining large propellers.

To ensure the machines could be exported smoothly, the people involved tampered with the declaration materials, disguising the advanced multi-axis equipment as lower-technology two-axis machine tools. Through false declarations, over two years, the Soviet Union successively obtained four nine-axis machine tools and later another four five-axis machine tools.

The deal initially escaped export regulatory authorities until 1987, when the matter was exposed by the U.S. media. The U.S. defense department believed that this batch of equipment might help the Soviet Union manufacture quieter submarine propellers, making it increasingly difficult for the U.S. military to track Soviet submarines.

When the news spread, the U.S. Congress exploded. Members of Congress not only swung hammers to smash radios but also called on American consumers to boycott Toshiba products. The Senate even passed a trade sanctions proposal against Toshiba by an overwhelming margin of 92 votes in favor and 5 against.

Toshiba's parent company was dragged into the uproar as well. To quell the anger, the company ran full-page apology ads in dozens of U.S. newspapers, and both the chairman and the president resigned. Then-Japanese Prime Minister Yasuhiro Nakasone publicly condemned the deal and even sent the Minister of International Trade and Industry to Washington to apologize.

This was the famous "Toshiba incident" of the Cold War era. The U.S. reaction at the time also shows just how much weight high-performance CNC machine tools carry in the defense industry.

Interestingly, what caused such a huge storm back then was an industrial giant weighing hundreds of tons. Today, however, machines with five-axis simultaneous control capability have been shrunk by Chinese teams to fit on a single workbench.

How on earth did they do it?

First, a fact that is easy to overlook: although a five-axis machine tool has only two more rotational directions than a three-axis one, its manufacturing difficulty increases by a large margin.

Take machining a metal impeller: the cutting tool has to reach between two curved blades, cutting while adjusting its angle.

The five axes must coordinate perfectly; the slightest deviation can ruin the part or even cause a direct tool crash.

The machine must also be stable enough. After all, it is cutting metal, with the tool spinning at high speed—even a slight vibration of the machine body can affect machining accuracy. The reason industrial machine tools of the past were so big and heavy—thick cast-iron beds and precision transmission mechanisms—was all to make them cut more accurately and more stably.

However, engineers began pondering early on: if you are only machining a few palm-sized parts, is it really necessary to build the machine tool so large?

In 2015, a husband-and-wife team of American engineers developed a desktop five-axis machine tool called Pocket NC.

This machine looks like a small metal box and can machine materials such as aluminum, brass, and plastic. Crowdfunding prices started at $3,500, and it ultimately attracted 280 backers.

But a new problem emerged—the machine had shrunk, but the difficulty of operating it did not shrink along with it.

To machine a part, drawing the 3D model is not enough. The material must be secured, the position measured, and the tool path and the angle of entry planned in advance. The machine may also need to be stopped mid-process to change tools and readjust.

You may have just wanted to make a watch case, but it turns out you first have to learn a whole set of CNC machining skills. For ordinary hobbyists, this barrier is still quite high.

And that is exactly the trouble that the Chinese teams crowding into this space in recent years want to solve. Among these companies are veteran players long engaged with industrial machine tools, as well as engineers who came out of consumer hardware companies such as DJI. The former are familiar with precision machining and know how to make machine tools cut accurately and stably; the latter excel at integrating complex technology into a single machine and at figuring out how to make it easy for users to operate.

These two kinds of experience happen to address the two major challenges of desktop five-axis machining: ensuring machining capability while lowering the operating barrier.

For example, in the past, the material position had to be measured manually before machining, and the machine had to be stopped mid-process to change tools. Today, some products have built laser scanning positioning and a six-tool-position automatic tool changer into the machine body; the machine can help determine the workpiece position and then automatically switch tools according to machining needs, saving a great deal of manual work.

The harder hurdle is programming. Machine tools cannot understand the 3D models designers draw. In the past, this work required operators to be proficient in professional CAM (computer-aided manufacturing) software, and learning it alone took considerable time. Today, some Chinese teams have begun developing companion software and are even trying to introduce AI to assist in generating tool paths and planning machining steps, so that users have less to do with complex machining instructions.

In fact, automatic positioning, automatic tool changing, and CAM software are not new inventions—large industrial machine tools have long used these technologies. What the Chinese teams are doing is combining the accumulated technology of industrial manufacturing with the product thinking of consumer electronics and packing it into a single machine.

More than a decade ago, when desktop 3D printers first appeared, they were also like half-finished products: you had to assemble them yourself, level them manually, and repeatedly test temperatures. After you bought one and took it home, the thing you printed most often was frequently the parts used to fix the machine itself.

Later, Chinese teams added features one by one—automatic leveling, vibration compensation, filament-runout detection, mobile apps—and only then did the machines go from usable to easy to use. People who know nothing about modeling can download a ready-made file and print figurines, board-game accessories, or even replacement plastic clips for broken household items.

The market responded quickly. In 2025, global shipments of entry-level 3D printers priced under $2,500 grew 26%, with Chinese manufacturers contributing more than 90%; at their peak, four Shenzhen companies—Creality, Bambu Lab, Anycubic, and Elegoo—once captured 94% of shipments in that price tier.

Laser engraving and cutting machines have gone through a similar transformation. The Chinese company xTool equipped its machines with camera positioning, enclosed housings, exhaust systems, and companion software, turning equipment once full of burnt smells and parameter tables into a personal creativity tool—in the first nine months of 2025, it accounted for about 47% of the global laser engraving and cutting market by GMV, with more than 400,000 connected devices.

3D printers, laser engravers, and desktop CNCs work on different principles, yet they share a similar development path: the core technology has existed for years, and Chinese companies, through software-hardware integration, gradually hand over to the machine steps that once required professionals.

Behind this lies another advantage: the supply chain that China's manufacturing industry has built up over many years.

For example, the "two-hour supply chain circle" in and around Shenzhen can assemble about 80% of the core components of a 3D printer; motors, guide rails, control boards, cameras, and sensors simultaneously serve phones, drones, robots, and smart home appliances.

For hardware companies, this means that when developing new products, they do not have to start everything from scratch. Whatever parts are needed, suppliers can be found nearby; if a design is not ideal, it can be re-prototyped and adjusted. Development cycles can be shortened, and the cost of trial and error reduced. Once a product sells in larger volumes, scaled production can further dilute costs, giving the machine a chance to be sold at a price more people can accept.

Once prices and operating barriers come down, the market changes too. Equipment that once only professional workshops would buy begins to appear in personal studios, home garages, and small startup teams. These new demands in turn push companies to keep improving their machines and developing software, making the equipment even easier to use.

Looking back, Chinese companies have walked this road more than once: using mature industrial technology and manufacturing supply chains to gradually turn niche, expensive, hard-to-operate equipment into products that more people can afford and use.

310 articles published on Huxiu. This content comes from the internet, original link; the views represent only the author and do not represent Huxiu's position. For copyright issues, please contact hezuo@huxiu.com, and we will verify and handle it promptly. Everyone who is changing the world and wants to change the world is on the Huxiu APP[ ]