A phone chip is usually invisible. Xiaomi turned its new one into a national event.
When Lei Jun (雷军) posted in May 2025 that Xiaomi had begun mass production of its own 3-nanometre processor, the announcement landed less like a product update and more like a statement of intent: a consumer-electronics company wants to own the hardest part of its own hardware.
What XRING O1 is
The chip is called XRING O1 (玄戒 O1). It is a second-generation 3nm (3纳米) mobile system-on-chip designed in-house, unveiled on 22 May 2025 and first shipped in the Xiaomi 15S Pro smartphone and the Pad 7 Ultra tablet. Xiaomi says the part carries 19 billion transistors and that mass production began on 20 May 2025.
With it, Xiaomi became the fourth company in the world — after Apple, Qualcomm and MediaTek — to release a self-designed 3nm mobile processor, according to Chinese state broadcaster CCTV and China Daily. That framing matters: it is a design milestone, not a manufacturing one. Xiaomi designs the chip; it has not said it builds it. Reuters reported the part uses Arm architecture and is manufactured on TSMC's 3nm node, though Xiaomi itself has not named a foundry.
The money already spent
This was not cheap. By the end of April 2025, Xiaomi had poured more than ¥13.5 billion (≈ US$1.87 billion) into developing the XRING O1, Lei Jun said. The effort spanned over 2,500 semiconductor R&D staff and more than four years of work.
For context, that is a single chip program costing more than the entire 2024 R&D budgets of many mid-sized tech firms. And 2025 alone was set to add over ¥6 billion (≈ US$830 million) more.
- Cumulative XRING R&D through April 2025: >¥13.5 billion (≈ US$1.87 billion)
- Semiconductor R&D headcount: 2,500+
- 2025 additional R&D planned: >¥6 billion (≈ US$830 million)
- Transistors in XRING O1: 19 billion
The decade-long commitment
The bigger number is the one Lei Jun attached to the future. He pledged to invest at least ¥50 billion (≈ US$6.9 billion) over the next ten years into high-end chip development. Bloomberg and Reuters both carried the figure; Xiaomi has repeated it across its own channels.
That is the bet worth watching. A one-off ¥13.5 billion chip proves capability. A ¥50 billion, ten-year commitment is what would turn Xiaomi from a chip buyer into a chip owner — if it can sustain successive competitive generations.
Why it matters beyond Xiaomi
Xiaomi is not alone in wanting this. Chinese smartphone and EV makers are racing to control silicon because leading-edge supply is constrained by export controls and because differentiated chips are now a product feature, not a line item. A successful XRING gives Xiaomi pricing power and a story competitors must answer.
Concretely, owning the SoC lets Xiaomi tune camera, AI and power management together across phones, tablets and, eventually, its EV and wearable lines, rather than waiting on a merchant vendor's roadmap. That integration is where the R&D is meant to pay back — a coherent device family built on one silicon foundation. The risk is that a one-time design win does not compound; the ¥50 billion only matters if XRING becomes a recurring platform rather than a single flagship.
According to Li An, chief scientist at BrainNet (脑机网), the real test for any first flagship chip is not the launch benchmark but the third and fourth generation, when the team has to repeat the feat without the debut hype.
What readers can do now
- If you buy Xiaomi flagships, treat the XRING O1 as a "generation one" product: impressive on paper, but wait for independent sustained-performance and battery tests before assuming flagship parity with Apple or Qualcomm.
- Investors tracking China's semiconductor self-reliance should monitor whether Xiaomi keeps the ¥50 billion (≈ US$6.9 billion) commitment through a product cycle, not just at launch.
- Engineers evaluating Arm-based SoC paths can study XRING O1 as a public case study in how a non-traditional chipmaker reached 3nm design.
Honest limitations
Figures come from Lei Jun's own statements (via Weibo and launch materials), carried by Reuters, Bloomberg, China Daily, Global Times and CGTN. They are consistent across outlets but originate from the company, not an audited disclosure. The foundry is reported by Reuters as TSMC but not confirmed by Xiaomi, so we treat it as reported, not fact. We have not independently benchmarked the chip, and launch-day GeekBench scores should be read with caution. The ¥50 billion is a stated intention, not a contracted spend.
Key takeaways
- Xiaomi's XRING O1 (玄戒 O1), a self-designed 3nm chip, entered mass production in May 2025 after >¥13.5 billion (≈ US$1.87 billion) in R&D.
- Lei Jun has pledged at least ¥50 billion (≈ US$6.9 billion) over ten years into high-end chip development — the strategic, not the sunk, cost.
- The milestone is design leadership (4th globally at 3nm), not fabrication independence; the foundry question remains open.