AI News AI资讯 1d ago Updated 1d ago 更新于 1天前 46

On the Eve of Listing: 35-Year-Old Tsinghua Graduate Pushes for HK IPO, Achieves Second Place in China's Smartphone Fingerprint Market 上市前夜|35岁清华学霸冲港股IPO,做出手机指纹中国第二

JiioV Technology has filed for a Hong Kong IPO, positioning itself as China’s second-largest smartphone biometric solution provider with a 14.5% market share. The company achieved a profitability inflection point in 2025, driven by a shift toward high-margin optical fingerprint sensors and a 50.1% three-year revenue CAGR. JiioV differentiates itself through a "perception interaction large model" that unifies multiple modalities (fingerprint, face, gesture), enabling expansion into smart glasses 极豪科技冲刺港股IPO,2025年营收4.69亿元,中国智能手机生物识别市场份额第二(14.5%),光学指纹传感器成为增长主引擎。 公司采用Fabless模式,通过算法优化将算力需求降低1000倍,实现高精度模型在边缘端的部署,光学指纹毛利率达45.8%。 团队源自旷视科技与豪威科技,拥有深厚的AI视觉与半导体产业链背景,正尝试通过“感知交互大模型”向智能眼镜和具身智能领域拓展。 财务显示经调整净利润转正,但IFRS口径仍亏损;客户高度集中,前五大客户占比近90%,单一最大客户占比超40%。

65
Hot 热度
70
Quality 质量
60
Impact 影响力

Analysis 深度分析

TL;DR

  • JiioV Technology has filed for a Hong Kong IPO, positioning itself as China’s second-largest smartphone biometric solution provider with a 14.5% market share.
  • The company achieved a profitability inflection point in 2025, driven by a shift toward high-margin optical fingerprint sensors and a 50.1% three-year revenue CAGR.
  • JiioV differentiates itself through a "perception interaction large model" that unifies multiple modalities (fingerprint, face, gesture), enabling expansion into smart glasses and embodied intelligence.
  • Despite strong growth, the company faces risks from high customer concentration (top client accounts for ~41% of revenue) and reliance on a supply chain dominated by major image sensor players like OmniVision.

Why It Matters

This case highlights the strategic pivot of AI hardware companies from single-sensor solutions to unified multimodal perception models, which is critical for scaling into emerging sectors like robotics and AR/VR. For investors and practitioners, it demonstrates how algorithmic efficiency (reducing compute needs by 1000x) can create significant moats in mature semiconductor markets, allowing smaller players to outpace giants like Goodix in growth rates.

Technical Details

  • Core Technology: Utilizes a proprietary "perception interaction large model" that integrates fingerprint, facial recognition, gesture, eye-tracking, and tactile data into a single architecture, reducing computational requirements by over 1000 times while maintaining precision.
  • Product Mix: Revenue is dominated by optical fingerprint sensors (60% of total, 45.8% gross margin) used in mid-to-high-end smartphones, while capacitive sensors (30.9% of total, 15% gross margin) serve as legacy volume drivers.
  • Business Model: Operates as a Fabless chip and algorithm company, selling integrated chip-algorithm combinations to module manufacturers rather than directly to OEMs.
  • New Verticals: Developing specific solutions for smart glasses (already validated by Xiaomi and Honor) and embodied intelligence (5mm ultra-thin sensors for robots, expected Q2 2026).

Industry Insight

  • Consolidation vs. Disruption: In mature biometric markets, growth is no longer driven by smartphone unit sales but by value-add features like under-display optics and multimodal integration; companies must innovate on the algorithm-hardware interface to capture margin.
  • Supply Chain Dynamics: High customer and supplier concentration poses significant risk; however, partnerships with ecosystem giants (e.g., OPPO, OmniVision) provide essential validation and manufacturing access for early-stage hardware startups.
  • Valuation Implications: The flat valuation between Pre-A (2022) and A round (2024) suggests the market priced in execution risks; the upcoming IPO will test whether the narrative of "multimodal AI for embodied intelligence" can justify a premium over traditional component suppliers.

TL;DR

  • 极豪科技冲刺港股IPO,2025年营收4.69亿元,中国智能手机生物识别市场份额第二(14.5%),光学指纹传感器成为增长主引擎。
  • 公司采用Fabless模式,通过算法优化将算力需求降低1000倍,实现高精度模型在边缘端的部署,光学指纹毛利率达45.8%。
  • 团队源自旷视科技与豪威科技,拥有深厚的AI视觉与半导体产业链背景,正尝试通过“感知交互大模型”向智能眼镜和具身智能领域拓展。
  • 财务显示经调整净利润转正,但IFRS口径仍亏损;客户高度集中,前五大客户占比近90%,单一最大客户占比超40%。

为什么值得看

极豪科技展示了AI算法如何重构传统硬件赛道,通过“算法定义硬件”实现降本增效,为AI落地边缘计算提供了商业闭环样本。同时,其从成熟手机市场向新兴具身智能领域的跨界尝试,反映了生物识别技术在多模态交互中的潜在延展价值。

技术解析

  • 核心技术与架构:极豪科技本质为Fabless芯片+算法公司,拥有感知交互大模型。其关键技术突破在于能在大幅保留模型精度的前提下,将算力需求降低1000倍以上,使复杂的多模态识别(指纹、人脸、手势等)能在资源受限的边缘端运行。
  • 产品线与收入结构:2025年90.9%收入来自智能手机解决方案。其中光学指纹传感器(屏下高阶方案)收入2.81亿元,毛利率45.8%,是主要利润来源;电容式指纹传感器(侧边/背面低阶方案)收入1.45亿元,毛利率仅15%且增长停滞。
  • 新业务拓展(NRE):技术服务(NRE)收入从2023年的247.9万暴增至2025年的3996.4万,毛利率高达62.3%。这部分收入主要来自智能眼镜和具身智能客户的前期定制开发,如小米、荣耀的智能眼镜及vivo Vision的手势方案。
  • 运营效率指标:2025年贸易应收周转仅19天,存货周转76天,应付账款46天,显示出较强的现金流管理和上游议价能力,经营活动现金流转正为1225万元。

行业启示

  • AI赋能传统硬件的边际效应:在成熟的生物识别赛道,单纯硬件堆叠难以获得高毛利,通过AI算法优化(如降低算力需求、提升识别精度)可以显著提升产品附加值和毛利率,是硬件公司转型的关键路径。
  • 多模态交互是具身智能的入口:极豪科技将指纹、人脸、手势、眼动等统一在感知交互大模型架构下,表明未来人机交互将从单一模态向多模态融合演进,具备统一底层算法能力的企业更容易横向扩展至机器人等新终端。
  • 供应链风险与集中度管理:作为上游芯片/算法提供商,极豪科技面临极高的客户集中度风险(单一客户占比超40%)。在IPO及后续发展中,如何分散客户结构、增强对终端品牌的直接影响力,将是决定其长期估值稳定性的关键因素。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Product Launch 产品发布 Funding 融资