Alibaba's T-Head files for HK IPO at up to $80B, rivaling Baidu's Kunlunxin
创业邦 · wechat · 2026-11-22
Alibaba's chip unit T-Head secretly filed for a Hong Kong IPO on Sept 11, with international banks valuing it at up to $80 billion; Baidu's Kunlunxin filed in January and confirmed its listing push on Nov 16.
Two different paths
- Baidu started FPGA exploration around 2010 (5,000+ units by 2015), launched Kunlun chips in 2018, and spun the unit out in 2021 at a 13B RMB valuation, now worth 21B RMB with Baidu holding 57.67%. It has won 1B+ RMB procurement orders from China Mobile.
- Alibaba formed T-Head in 2018 by merging CentecMicro-acquired Hanguang teams; its Hanguang 800 cut image-search processing from 1 hour to 5 minutes. Kept as a wholly-owned subsidiary, it now generates 10B RMB annualized revenue, shipped 560,000 Zhenwu-series chips, serves 650+ external customers, and powers Moonshot's KimiK3 via the M890 supernode.
Why the same business
Alibaba pledged 380B RMB over three years for cloud/AI infrastructure; Baidu's ad revenue fell 22% YoY while AI cloud infra revenue grew 94%. In-house chips protect cloud margins. China shipped 2M+ AI accelerators in H1, led domestically by Huawei, T-Head, Cambricon, and Kunlunxin. Post-IPO, the two will compete for the same customers, engineers, and investors.
More from Infra
- An AI bubble fable: from free cash flow to private credit to annuity blowup — Svwang1 · 2026-11-22
- One RX 7900 XTX runs 524k context at 767 tok/s, 3x faster than llama.cpp — W61k3r · 2026-11-22
- How GPUs organize work: kernel launches, thread blocks and scheduling explained — blaizedsouza · 2026-11-22
- Can llama.cpp RPC span a Mac mini, Mac Studio and dual RTX 3090s to run a 120B model? — ruffus_or · 2026-11-22
- OpenAI's GPT rewrites AMD flagship GPU RTL, gaining up to 14% PPA in 5 months — 新智元 · 2026-11-22
- LBNL report: data center load growth did not drive up residential electricity prices (2019-2025) — Eliv_nurotic · 2026-11-22