Editor’s note: this piece is about a Chinese motorcycle manufacturer and what it says about how Chinese factories get better. If you buy machined parts from China, that’s the question underneath this whole story.
1. Let’s be precise: this is September, and they are not “champions again”
On 28 March 2026, at the Portimão circuit in Portugal, French rider Valentin Debise crossed the line on a Zhang Xue Motorcycles 820RR-RS, 3.685 seconds ahead of second place. He won again the following day.
It was the first time a Chinese motorcycle brand had won in the WorldSSP middleweight class of the Superbike World Championship — breaking a decades-long hold by European and Japanese manufacturers on that class.
Where things stand now: at the French round on 5–6 September, at the Nevers Magny-Cours circuit, Debise took the first pole position of his WorldSSP career and finished third in both races, standing on the podium twice in one weekend. He sits second in the riders’ standings on 241 points. Zhang Xue Motorcycles sits third in the manufacturers’ standings on 243 points, behind Yamaha (370) and Ducati (298).
So the accurate statement is not “they won again”. It’s nine rounds into the season, still solidly in the leading group.
And that distinction matters.
A win can be a burst. Staying in the top three takes a system. One victory can be luck. Being third after nine rounds means the bike, the team and the supply chain have genuinely got something right.
That’s what this article is about — what they got right, and what it means for Chinese manufacturing generally.
2. Two numbers explain more than the trophy does
The first number: 0.02 mm
The company that makes the crankshaft is Tianrun Industry, in Shandong. They moved their crankshaft tolerance from 5 si to 3 si.
“Si” is Chinese shop-floor shorthand: 1 si = 0.01 mm. So the improvement was from 0.05 mm to 0.03 mm — a shift of 0.02 mm.
How small is that? A human hair is roughly 0.07 mm. The step they took is less than a third of a hair’s width.
Looks negligible, doesn’t it?
But this 819cc three-cylinder engine has a redline of 15,250 rpm and has been pushed to 16,000 in testing. These parts have to survive 1,800 hours on the test bench — roughly seven years of riding for an average owner.
At that engine speed, 0.02 mm is what decides whether the thing survives all 1,800 hours.
The second number: 20 years
Zhang Xue was born in 1987 in Mayang, Hunan province. He left school at 14 and went to work as an apprentice in a motorcycle repair shop.
In 2006, aged 19, he wanted a chance to show his riding on television. So he rode a battered old motorcycle through a rainstorm for over 100 kilometres, chasing the TV crew’s car. They sped up several times to shake him off. He kept following. When they finally stopped, he walked up to the window and said:
“I’m not afraid of hardship. I’m just afraid you won’t give me a chance.”
He went on to race professionally, then moved into building motorcycles after an injury. In 2013 he went to Chongqing with 20,000 yuan to start a company. In 2017 he co-founded Kove, a Chongqing motorcycle brand, and later led a team to the finish of the Dakar Rally — the first Chinese motorcycle manufacturer ever to complete it.
In 2024 he walked away from all his equity in Kove, took nothing with him, said “I’m going to chase my stars” and started again. In March 2026, he stood on top of the world championship podium.
Between the teenager chasing a car through the rain and that trophy: twenty years.
3. The third number is the real watershed: the 0.1% he volunteered
After the championship win, someone asked Zhang Xue on camera what the localisation rate was.
He didn’t check a spreadsheet or call an assistant. He crouched down beside the winning bike and counted — screw by screw, shim by shim. A few minutes later he stood up and gave a number: 99.9%.
The remaining 0.1% was two parts most people would never look twice at: the bearing shells and the timing chain inside the engine. Both from Japan.
He could have stopped at 99.9%. The whole internet was celebrating. Nobody was going to audit that 0.1%. Instead he volunteered:
“It’s not that we can’t make them. It’s that under extreme conditions, the domestic parts are still slightly less consistent in tensile and wear resistance. For the racing version, to win, we used imported parts.”
I think that sentence is the single most instructive thing in the entire story.
Chinese entrepreneurs generally carry one belief: externally, you must look like you can do everything. Admitting a weakness feels like losing face — like the customer will think less of you.
Real business works the other way round.
A supplier who will tell you “we can hold this tolerance on this part; that other part we can’t do, go talk to these people” is far more reliable than one who thumps the table and says “no problem” to everything.
Because the first one is working from a clear internal ledger — he knows exactly where his capability ends. And a company that knows where its capability ends is a company that can deliver.
Confidence isn’t “I can do everything”. Confidence is “I know exactly what I’m good at, and I’m not afraid to say what I’m not.”
That may be the lesson Chinese manufacturing most needs to learn: moving from “never lose face” to “dare to define yourself”.
4. The factories that don’t hold press conferences are the real foundation
The championship bike is more than 90% domestically sourced. Of 39 core suppliers, 34 are Chinese, and half of those are in Chongqing.
You have probably never heard of most of those 34:
| Company | What they supply | What they achieved |
|---|---|---|
| Tianrun Industry (Shandong) | Crankshaft, con rods | Crankshaft tolerance from 0.05 mm to 0.03 mm; 1,800-hour bench test |
| Zhongyuan Neipei (Henan) | Cylinder liners | Sealing integrity under high temperature and pressure |
| Jihua Laboratory (Guangdong) | TiAl valves | 30% lighter than conventional titanium, rated above 850°C |
| Keboda | ECU | Independent engine management |
| CETC (Chongqing) | 6-axis IMU | Cornering ABS and traction control algorithms |
| Ningbo Saifu | ABS system | Brake force adjusted within 0.01 seconds |
Not one of them held a launch event. Not one chased the news cycle. Not one put out a slogan.
They all did the same thing: they pushed one specific, measurable technical number a little further forward.
Tianrun’s 0.02 mm is the purest example of that logic.
Progress in Chinese manufacturing is often not a leap “from 0 to 1” — it’s a climb “from 5 si to 3 si”. The first makes headlines. The second decides whether you survive.
And what does that climb require?
Twenty years of accumulation that nobody sees.
5. Long-term thinking isn’t a slogan. It’s arithmetic.
“Long-termism” has been used so often it’s nearly become a platitude.
But in this case, it’s calculable:
In 2025, Zhang Xue Motorcycles spent 69.58 million yuan on R&D — 9.33% of revenue. For context, the industry average is 3–5%. In other words, he spent two to three times what his competitors did.
The cost: 75 million yuan in revenue, a book loss of 22.78 million yuan, and 25,000 motorcycles sold for the year.
A company founded in 2024, not yet profitable, putting nearly a tenth of its income into R&D and track testing.
That is what long-term thinking actually looks like. Not “I am patient” — but “I am willing to take a visible loss in exchange for an invisible future.”
Then came March 2026, and the championship. Orders came through the gates — 20,000 in ten days. The 2026 R&D budget rose to 135 million yuan.
But note the sequence: he lost 22.78 million yuan first. Then he won. You can’t reverse the order.
A lot of companies try to copy long-termism as “I will persevere”, and miss the part that matters: “I will keep committing quantifiable resources while the return is invisible.” The first is an attitude. The second is a capability.
Zhang Xue once said something I’ve kept with me:
“You don’t do a thing for the result. You do it because you love it — and then maybe the result turns out different.”
That reads like a greeting card. But put it next to 69.58 million yuan of R&D spend and it means something else: love isn’t a feeling. Love is being willing to keep paying for something that hasn’t paid you back yet.
6. Speed comes from the cluster. Consistency is built by hand.
Zhang Xue Motorcycles was founded in 2024 and was on a world-class grid within two and a half years. How?
Chongqing’s one-hour industrial circle: 51 large-scale motorcycle manufacturers, more than 410 large-scale component suppliers, and over 80% local sourcing for combustion motorcycles. The local saying is that you can source every part of a motorcycle within 40 kilometres.
At the 2026 Summer Davos, that circle was named alongside the Yangtze River Delta’s one-hour electronics innovation circle and the half-hour supplier radius of Shenzhen’s robotics valley as representative Chinese industrial clusters.
Clusters solve the speed problem. A new model that takes 24 months from design to production elsewhere can be compressed to 12–15 months in Chongqing. A founder never has to leave the circle to model, prototype, test and manufacture.
But a cluster cannot solve the other problem: batch consistency.
Producing one good part in prototyping, and producing five thousand good parts in a row, are entirely different trades. The championship bike finished race after race not because any single part was exceptional, but because the parts from dozens of suppliers — over a thousand components — each held their own stability under extreme conditions.
An engine that runs 1,800 hours at 15,000 rpm is, at bottom, saying this: the quality variation across this one batch, across a thousand parts, has been held inside a very narrow band.
That is what “internal strength” actually means. It isn’t flashy. It just doesn’t fluctuate.
7. So what should a Chinese company actually work on?
Stripping the case down to manufacturing practice, I’d say four things:
First, turn self-knowledge into something you can deliver.
Zhang Xue could say 99.9% out loud because he had a clear ledger in his head: which parts were fine, which fell short, and by how much. Every company should have that ledger — which processes are genuine advantages, which are merely “we can do it”, and which still have to be outsourced. Companies that will lay that ledger open are the ones customers trust.
Second, replace feeling with data.
The credibility of 0.03 mm comes from 1,800 hours of bench testing. Machining is the same: CPK, SPC and CMM data are not decoration for the customer. They are the only proof that what you do is stable. Traceable process data is part of the product.
Third, treat R&D as a cost of goods, not an expense.
Spending 69.58 million to book a 22.78 million loss makes for an ugly set of accounts. But that spending is what produced the 2026 championship and the order book. If all of a company’s value shows up as an hourly rate on a quote, it will eventually be undercut by somewhere cheaper. The real moat is the processes nobody else can quote on — or can quote on but can’t hold a stable batch of.
Fourth, accept that some things simply take time.
Apprentice at 14. Chasing a car through the rain at 19. Twenty thousand yuan to start a business at 26. Dakar in between. Walking away from everything at 40 to start again. Champion at 42. Twenty years in between.
Saying “manufacturing has no shortcuts” gets old fast. What makes this case different is that he chose the slowest possible road in an industry that looks like it rewards cheap volume — and won.
And that, in its way, is the story of Chinese manufacturing over these decades. We didn’t get here by shortcuts. We got here because a generation ground 5 si down to 3 si, and 0.05 down to 0.03 — one hundredth of a millimetre at a time.
So Chinese companies can afford a bit more confidence in themselves.
Not the “we can do anything” kind. The kind that says: “I know exactly how far I’ve come, and I know where I’m going next.”
8. One last thing: the brakes
Having said all that, the boundary needs stating clearly.
This success has strong case-specific features. The founder had twenty years of industry depth. Chongqing has the densest motorcycle cluster in China. And China’s supply chain in 2026 is more complete across more categories than at any point in its history. All three conditions lining up at once is not easily repeatable.
And in absolute scale, the company is still a long way behind established players like CFMoto and Qianjiang.
So the conclusion is not “copy this and you win”. It’s this:
In an industry where everyone competes on price, making the technical numbers solid, stating your boundaries clearly, keeping your data, and treating R&D as an investment rather than an overhead — that road has now been validated by a twenty-year case.
It’s slow. But it doesn’t need luck.
In closing
On the podium stands the rider. Holding the trophy is the founder.
But what actually pushed the bike up there was people in factories that don’t hold press conferences, moving precision forward by one si, day after day.
The thickest layer of Chinese manufacturing has never been under the spotlight.
Its confidence comes from twenty years, from 0.02 mm, and from all those unseen days when nobody stopped.
Sources
Race results and standings: Xinhua, CCTV Sports, China News Service (6–7 September 2026). Supplier list and localisation rate: Zhang Xue’s public statements and associated reporting. R&D spend, revenue and unit sales: Shanghai Securities News. Chongqing cluster data: Chongqing Municipal Government, Economic Daily. Biographical details: Xinhua, People’s Daily, 21st Century Business Herald.
About Eternal CNC
Xinyongheng (Shenzhen) Precision Industrial Co., Ltd. sits on the same supply chain. 23 years in the manufacturing industry, 15 years in CNC precision machining, 30 CNC machines, milling to ±0.005 mm — our work is also about pushing parts from “we can make it” to “we can make it consistently”.
In robotics, we machine and prototype precision parts for joint modules, reducer components, end effectors and sensor brackets, from single prototypes to low-volume production runs.
If you have a part that’s giving you trouble, send it over. Our engineers run a DFM review first and come back with a quote within 2 hours. Prototypes in 3–7 days, production in 15–30 days, no minimum order quantity.