Table of Contents
Published: 2026-09-18 · Category: RISC-V MCU · Reading time: ~5 min · Status: DRAFT
The statement and where it was made
At the 6th Smart Vehicle Chip Ecosystem Conference on 16 September 2026, Infineon's Liu Lin (Senior Manager, Automotive Marketing; RISC-V ecosystem lead for Greater China) stated that Infineon's next-generation automotive MCU products will use the RISC-V architecture across the board — quoted as going "all in" on RISC-V.
The claim is being carried by Chinese automotive trade press (Gasgoo, ifeng Auto, Autohome) that all reproduce the same talk. No equivalent English-language Infineon press release was found; treat this as a conference statement reported by trade media, not a product announcement with a datasheet behind it.
Today's portfolio is two architectures; the next one is one
Infineon's current MCU line-up is explicitly split:
| Segment | Family | Core | Applications cited |
|---|---|---|---|
| Low / mid-end | TRAVEO T2G | Arm | Body control, HMI |
| Mid / high-end | AURIX (TC2 / TC3 / TC4) | TriCore (in-house) | Domain controllers, powertrain, BMS |
The next generation consolidates both onto RISC-V. That is the part worth pausing on: this is not a RISC-V variant added beside an incumbent architecture, it is a consolidation of two existing architectures into one, including the replacement of Infineon's own proprietary TriCore ISA at the high end where functional-safety requirements are heaviest.
Scale context given in the same talk: Infineon's automotive electronics business was 50% of FY2025 revenue, Greater China is 38% of global business, MCU share moved from 32.1% (2025) to 36%, and per TechInsights (April 2026) Infineon has been the largest automotive semiconductor supplier since 2019.
The four stated reasons
Liu gave four rationales, and they are the ones a procurement or architecture team should interrogate:
- Flexibility — one unified architecture covering low-end through high-end.
- Open standard — no dependency on a single IP supplier.
- Modularity / extensibility — faster iteration.
- Ecosystem — existing collaborative work.
Reason 2 is the one with real weight for a Tier-1 doing multi-decade platform planning: an open ISA with multiple independent core suppliers is a multi-source supply argument, not an ideology argument. It is also the same reason a supplier can commit to long-term availability without a single licensor's roadmap constraining it.
Toolchain and OS work already done
The useful content in the talk is the ecosystem checklist, because it is concrete and dated:
- 2024 — baseline toolchain in place: compiler, debugger, virtual prototype.
- 2025–2026 — moved up to the OS layer. AUTOSAR adaptation with Vector, Neusoft Reach (东软睿驰) and iSoft (普华); open-source side working with FreeRTOS, Zephyr and NuttX.
- March 2026 — DRIVECORE™ RISC-V virtual prototype pre-integrated evaluation package released, requestable from Infineon's site; some domestic OEMs and Tier-1s reportedly using it for early evaluation.
- Planned — a cloud-based MCU evaluation platform for software-defined-vehicle development.
For an engineering team, the March 2026 virtual prototype is the actionable item: it means architecture evaluation can start before silicon exists.
Standardisation
Two standardisation threads were cited:
- CACC (China Automotive Chip Standards, Testing and Certification Alliance) formed an automotive RISC-V working group in 2025; Infineon joined as a core member.
- The 《汽车RISC-V芯片标准化需求研究报告》 (Automotive RISC-V Chip Standardisation Requirements Research Report) was published July 2026, with Infineon contributing to drafting.
Note the regional split: Infineon is simultaneously inside RVI's AUTO-SIG work internationally and the CACC group in China. Fragmentation between those two tracks remains an open risk for anyone shipping globally.
What is not disclosed
This is a strategy statement, not a product launch. The following are absent from all coverage found and must not be inferred:
- No part numbers, no core names, no core counts, no clock speeds.
- No process node, no memory sizes, no package or pin-out.
- No functional-safety target (ASIL-B / ASIL-D) for the RISC-V generation, and no certification timeline.
- No sampling or production dates.
- No statement on TriCore's future beyond "next generation goes to RISC-V" — the AURIX TC2/TC3/TC4 installed base and its support horizon were not addressed.
- No migration path for existing AURIX or TRAVEO users.
Anyone quoting "Infineon is moving to RISC-V" without these caveats is overstating what was said.
Engineering takeaway
The signal is direction and commitment, not specifications. A supplier with roughly a third of the automotive MCU market saying its next generation is single-architecture RISC-V is meaningful for long-horizon platform planning — particularly for teams who care about multi-source availability and not being locked to one IP licensor.
The practical next step is the DRIVECORE RISC-V virtual prototype. Everything else — safety level, timing, performance, migration — is still unanswered, and until a numbered part with a safety manual exists, this should be treated as roadmap intelligence rather than a selection input.
Sources
- Gasgoo — "英飞凌刘琳:All in RISC-V,下一代汽车MCU全面转向开放架构", 16 September 2026 15:32 — https://i.gasgoo.com/news/70472188.html
- Gasgoo — "圆满落幕丨2026第六届智能汽车芯片生态大会", 17 September 2026 — https://auto.gasgoo.com/news/202609/17I70472152C106.shtml
- ifeng Auto / 网通社 — "英飞凌宣布下一代汽车MCU全面转向RISC-V架构", 16 September 2026 — https://auto.ifeng.com/c/8wT6kI7MI8c
- Autohome 车家号 — same talk, September 2026 — https://chejiahao.autohome.com.cn/info/26465989
- 同花顺财经 — "英飞凌宣布下一代汽车MCU全面转向RISC-V架构", 16 September 2026 — https://news.10jqka.com.cn/20260916/c679965194.shtml
Verification notes
- All figures (FY2025 automotive = 50% of revenue; Greater China = 38%; MCU share 32.1% → 36%; TechInsights April 2026 ranking; TRAVEO T2G = Arm; AURIX TC2/TC3/TC4 = TriCore; 2024 toolchain; 2025–2026 AUTOSAR with Vector / 东软睿驰 / 普华; FreeRTOS / Zephyr / NuttX; DRIVECORE March 2026; CACC working group 2025; report published July 2026) are quoted from the trade-press reports of the 16 September 2026 talk. Nothing was inferred.
- Single-source cluster: Gasgoo, ifeng Auto, Autohome, 10jqka and chejiahao all report the same conference talk. Numbers are consistent across them, but they are not independent verification. No Infineon English-language press release or official statement was found — this gap is stated in the body.
- Market-share figures (32.1% → 36%) are Infineon's own citation of TechInsights, relayed by trade press; the underlying April 2026 TechInsights report was not retrieved.
- No part number, core IP, frequency, process node, safety level or date exists in any source. These are listed as not disclosed rather than estimated.
- The talk was delivered in Greater China by Infineon's Greater China RISC-V ecosystem lead. Whether it represents a global product decision or a regional statement is not stated by any source; the draft does not claim it is global.
- Related but distinct from the Hot Chips 2026 Infineon talk on zonal E/E architecture already covered on this site. That talk covered architecture direction for zone controllers; this one is an explicit portfolio-level architecture commitment. The draft frames them as separate events.
- House style applied: framed as open-standard / multi-source-supply / long-term-availability reasoning. The Chinese sources use domestic-industry framing; it is not carried into this draft.