📰 DigiTimes 实时DigiTimes RealtimeDoubao · 实盘不做把握 5/10不做原因:证据不足
北京时间·🕒 发布 09-14 12:55:00时区推断·⟳ 抓取 09-14 13:03:07·⚡ 决策 09-15 13:23:42 +1d00h·42.1 天前
鸿腾精密(FIT)Tech Day称800G/1.6T光模块获全球服务器品牌设计定案并进入商用验证部署，向电源与液冷扩展
这是FIT Tech Day的高管营销性表态：核心'800G/1.6T获设计定案、进入商用验证'与2026年6月纪要已披露的'1.6T硅光H2验证、ELSFP H2产生收入'基本一致，无新客户、无订单金额、无指引上修；对约377亿港元市值公司属软催化。消息次日(9/15)股价不涨反跌约2.1%，处52周低点附近的下行通道，市场未给正反馈，缺乏可交易edge。
📄 原文DigiTimes Realtime↗ 原文
中文網 繁體 简体 SUBSCRIBE My account Sign out SUBSCRIBE Trending Apple Huawei Nvidia AI Search Scaling AI Glasses Inference E…
不做 鸿腾精密 FIT（06088.HK）
Tech Day营销性软催化，1.6T设计定案无客户/金额/指引增量且与6月旧纪要重合，对约377亿港元市值缺乏可量化edge；股价处下行通道、消息次日反跌2.1%无承接，不做多；无利空不做空。
展开详情(成交明细 / 研究档案 / 原文全文 / 决策记录链)
研究档案
结论摘要
DigiTimes报道FIT Tech Day重申1.6T光模块获设计定案并扩展电源/液冷，但无客户、金额或指引增量，且与6月已有叙事重复、股价处下行通道，属超大市值软催化，不做。
🔎 新鲜度:最早出现 2026-09-14 · 扩散 low · price-in 风险 medium ↗
关键理由
- 若做多个股，核心逻辑是1.6T设计定案与机架级信号-电源-液冷平台化，但该信息已在6月纪要中披露、本次无量化新增
主要风险
- 报道为Tech Day公关性访谈，未点名客户与订单规模，设计定案到规模收入仍有验证与份额不确定性
- 股价距52周高点约-52%、近13周持续下行，缺乏右侧承接，弱势股“低位”不构成做多理由
反向证据
- 做多前查反向 · 来源 kb_search · 日期 2026-06-26 · 影响 weakens
岗底斯6月纪要已披露1.6T硅光2026H2验证、ELSFP 2026H2产生收入、AI业务2026占比约25%，本次Tech Day表态基本是旧叙事重申，无新增惊喜
检索:FIT 鸿腾 1.6T 硅光 验证 收入 ELSFP 2026H2 纪要 原文 ↗ - 做多前查反向 · 来源 query_modeling_kb · 日期 2026-09-15 · 影响 weakens
行情显示6088.HK距52周高点约-51.6%、近13周指数化轨迹持续下行，消息次日仍跌约2.1%，无右侧资金承接
检索:6088.HK 近期股价 走势 利空 下调
先例如何影响了这次决策
对本次决策的影响:因软催化+无增量+股价下行，直接导向AVOID，未据弱先例设定tp/sl
方向:无可对照的同主体Tech Day/软催化对齐先例；本条为无量化增量的营销性表态，方向参考价值不足以支撑ENTER
涨幅量级:缺少同市场、同相位、首发锚干净且含金额/指引上修的对齐样本，无法据此校准止盈空间
形态:个股市场反应(次日-2.1%)与下行轨迹提示此类Tech Day叙事缺乏短线驱动
工具执行结果 · 成功 3 · 降级 0 · 失败 0 · 取消 0
- search_processed_events 成功 stock_tools HTTP · 14 ms
{"result": "search_processed_events:agent query=`fit 800vdc liquid cooling uqd cpo npo connector ai rack | foxconn interconnect technology fit 1.6t optical transceiver design win tech day | 鸿腾精密 fit 6088 1.6t 光模块 设计定案 技术日`；结构化索引覆盖完整 72 小时；旧记录兼容扫描近 0.0/72 小时、0 条本臂决策(完整)，返回 0 条候选。未命中只表示这些 agent query - stock_quote 成功 stock_tools HTTP · 291 ms
{ - kb_search 成功 stock_tools HTTP · 3,217 ms
{
原文全文
展开原文 · 7,167 字
中文網
繁體
简体
SUBSCRIBE
My account
Sign out
SUBSCRIBE
Trending Apple
Huawei
Nvidia
AI Search
Scaling AI Glasses
Inference Era AI
Advanced Packaging
Trending
SUBSCRIBE
CONNECT WITH US
Sign out
REALTIME NEWS
Corning sees CPO and glass core as new semiconductor battlegrounds
Tomorrow's Headlines
Sep 14, 18:47
China optical transceivers avoid US FCC curbs as NPO and CPO compete in high-speed links
Tomorrow's Headlines
Sep 14, 18:46
China's AIDC token costs hinge on access to advanced GPUs
Tomorrow's Headlines
Sep 14, 18:46
Home
Tech
ICT
Foxconn subsidiary broadens AI interconnect role as 1.6T enters deployment
Ninelu Tu
, Taipei
Sep 14, 2026, 12:55
0
Foxconn Interconnect Technology (FIT) senior director of US business development, Kai-pin Lin. Credit: FIT
AI data centers are moving decisively toward 1.6T connectivity, with 800G now mainstream in AI clusters and 400G remaining common in general cloud and enterprise deployments.
Kai-pin Lin, senior director of US business development at Foxconn Interconnect Technology (FIT), said the company's 800G and 1.6T optical transceivers have secured design wins with leading global server brands and entered commercial platform validation and deployment.
The transition is accelerating in 2026 as 224G SerDes platforms move into broader adoption. GPU-based platforms are progressing faster, with new flagship AI systems beginning to incorporate 1.6T into scale-out networks while evaluating CPO at the switch level.
Hyperscalers' in-house ASIC platforms, by contrast, still rely mainly on 800G and are expected to migrate more gradually to 1.6T, depending on network topology, switch architecture, optical circuit switching (OCS), and data-center refresh cycles.
The direction of travel is clear: the key difference is no longer whether data centers will move from 800G to 1.6T, but when they will upgrade and which network architecture they will use to absorb the higher bandwidth.
AI rack design forces signal, power and cooling to converge
Lin said FIT sees high-speed signals, power and cooling as interdependent elements in AI rack design, not separate component categories.
As AI servers move toward full-rack systems, high-speed transmission, high-power delivery and liquid cooling must coexist within limited space. That creates trade-offs across signal integrity, high-voltage and high-current power delivery, piping, service access and mechanical design.
Suppliers focused on a single component may therefore have less influence over platform specifications. FIT has expanded from high-speed connectors and cables into high-voltage power connectors and liquid-cooling UQD quick disconnects, extending its role into power and thermal management.
Foxconn's capabilities across boards, systems and full-rack manufacturing also give FIT earlier visibility into customer platform requirements. Lin said future AI rack design will increasingly depend on suppliers' ability to balance signal, power, thermal and mechanical constraints rather than optimize individual components in isolation.
FIT reframes CPO and NPO around the interface layer
Optical communications remain central to FIT's expansion, but its focus extends beyond optical engines to the interfaces linking chips, optical engines, modules and systems.
Lin said pluggable optics, NPO and CPO will still require detachable, serviceable and potentially blind-mate fiber interfaces. Even as optical engines move closer to switch chips or become co-packaged, external fiber connections still need high-density routing, precision alignment and maintainability.
CPO and NPO therefore do not eliminate connector demand; they change its location and raise technical requirements. FIT plans to extend its capabilities in precision tooling, micro-molding, stamping and automated assembly into MPO, high-density optical connectivity, fiber array units (FAUs), and future CPO and NPO interfaces.
As optical communications move from laboratory and low-volume use into large-scale data-center deployment, the competitive focus will shift beyond transmission performance toward connection accuracy, reliability, serviceability and volume manufacturing.
FIT moves from components to rack-level AI design
FIT Tech Day highlighted the company's broader portfolio across high-speed interconnects, power and cooling. Lin said its 800G and 1.6T optical transceivers have secured design wins with leading global server brands and are moving into commercial deployment.
FIT is also showcasing 224G interconnects, 800VDC connectors, liquid-cooling UQD products and high-density fiber connectivity for CPO and NPO, targeting rising AI data-center requirements for bandwidth, power density and cooling.
Supply-chain sources said optical communications are increasingly converging with copper cables, high-speed connectors, fiber interfaces, power and liquid cooling at the rack level.
For suppliers, the key shift is from selling individual components to participating earlier in next-generation AI platform design. Companies able to combine signal, power, cooling, and mechanical design into manufacturable system-level solutions will be better positioned as rack architectures evolve.
Article
translated by Levi Li and edited by Jerry Chen
AI Search
Related stories
AI data centers push liquid cooling into core infrastructure
Data center power and cooling overhaul will reshape global AI infrastructure
Delta targets next-gen AI data centers with 800V DC power and liquid cooling solutions
Analysis: AI server power shifts to rack-level architectures as Delta, Liteon take lead
OCP Summit: AI data centers embrace high-voltage power and liquid cooling
Show more
Categories
IT + CE Server, IPC, cloud computing, IoT
Tags
cooling CPO data design FIT switch
Share this article
Other links
Print
Related topic
Global supply chain: Mobile, telecom, computing
BIZ FOCUS
Sep 14, 08:00
Hauman Unveils Micro LED Testing Breakthrough for AI Optical Interconnects
Monday 14 September 2026
Scaling Photonics: Assembly, Test and Fiber Automation at SEMICON Taiwan
Friday 11 September 2026
DAS Expands in Taiwan and Unveils World First at SEMICON
Friday 11 September 2026
Toward Technologies Redefines Relay Solutions at SEMICON Taiwan 2026
MOST-READ
7 DAYS NEWS
Google's v10 TPU opens door to AMD without sidelining Broadcom, MediaTek
Update: ABF substrate maker Unimicron searched in origin-labeling probe
Qualcomm's HBC play pulls Samsung and SK hynix into a new race beyond HBM
SK keyfoundry expands 8-inch capacity as China demand strains supply
Samsung, SK Hynix step up China NAND upgrades as US tool risks grow
Column: When digital trust breaks — quantum risk rewrites supply chain resilience
Wuliangye skips carmaking, turns liquor waste into silicon-carbon anodes
The 1,000-watt problem: a TSMC veteran on the brute force forging Nvidia's AI chips
eCloudvalley expands enterprise AI Agent portfolio
CXMT rides AI boom to 80% operating margin
Full list
Members only
×
Sorry, the page you are trying to open is available only for our paid subscribers.
Please login to read more
New users, please register first
Notification
×
繁體
简体
SUBSCRIBE
My account
Sign out
SUBSCRIBE
Trending Apple
Huawei
Nvidia
AI Search
Scaling AI Glasses
Inference Era AI
Advanced Packaging
Trending
SUBSCRIBE
CONNECT WITH US
Sign out
REALTIME NEWS
Corning sees CPO and glass core as new semiconductor battlegrounds
Tomorrow's Headlines
Sep 14, 18:47
China optical transceivers avoid US FCC curbs as NPO and CPO compete in high-speed links
Tomorrow's Headlines
Sep 14, 18:46
China's AIDC token costs hinge on access to advanced GPUs
Tomorrow's Headlines
Sep 14, 18:46
Home
Tech
ICT
Foxconn subsidiary broadens AI interconnect role as 1.6T enters deployment
Ninelu Tu
, Taipei
Sep 14, 2026, 12:55
0
Foxconn Interconnect Technology (FIT) senior director of US business development, Kai-pin Lin. Credit: FIT
AI data centers are moving decisively toward 1.6T connectivity, with 800G now mainstream in AI clusters and 400G remaining common in general cloud and enterprise deployments.
Kai-pin Lin, senior director of US business development at Foxconn Interconnect Technology (FIT), said the company's 800G and 1.6T optical transceivers have secured design wins with leading global server brands and entered commercial platform validation and deployment.
The transition is accelerating in 2026 as 224G SerDes platforms move into broader adoption. GPU-based platforms are progressing faster, with new flagship AI systems beginning to incorporate 1.6T into scale-out networks while evaluating CPO at the switch level.
Hyperscalers' in-house ASIC platforms, by contrast, still rely mainly on 800G and are expected to migrate more gradually to 1.6T, depending on network topology, switch architecture, optical circuit switching (OCS), and data-center refresh cycles.
The direction of travel is clear: the key difference is no longer whether data centers will move from 800G to 1.6T, but when they will upgrade and which network architecture they will use to absorb the higher bandwidth.
AI rack design forces signal, power and cooling to converge
Lin said FIT sees high-speed signals, power and cooling as interdependent elements in AI rack design, not separate component categories.
As AI servers move toward full-rack systems, high-speed transmission, high-power delivery and liquid cooling must coexist within limited space. That creates trade-offs across signal integrity, high-voltage and high-current power delivery, piping, service access and mechanical design.
Suppliers focused on a single component may therefore have less influence over platform specifications. FIT has expanded from high-speed connectors and cables into high-voltage power connectors and liquid-cooling UQD quick disconnects, extending its role into power and thermal management.
Foxconn's capabilities across boards, systems and full-rack manufacturing also give FIT earlier visibility into customer platform requirements. Lin said future AI rack design will increasingly depend on suppliers' ability to balance signal, power, thermal and mechanical constraints rather than optimize individual components in isolation.
FIT reframes CPO and NPO around the interface layer
Optical communications remain central to FIT's expansion, but its focus extends beyond optical engines to the interfaces linking chips, optical engines, modules and systems.
Lin said pluggable optics, NPO and CPO will still require detachable, serviceable and potentially blind-mate fiber interfaces. Even as optical engines move closer to switch chips or become co-packaged, external fiber connections still need high-density routing, precision alignment and maintainability.
CPO and NPO therefore do not eliminate connector demand; they change its location and raise technical requirements. FIT plans to extend its capabilities in precision tooling, micro-molding, stamping and automated assembly into MPO, high-density optical connectivity, fiber array units (FAUs), and future CPO and NPO interfaces.
As optical communications move from laboratory and low-volume use into large-scale data-center deployment, the competitive focus will shift beyond transmission performance toward connection accuracy, reliability, serviceability and volume manufacturing.
FIT moves from components to rack-level AI design
FIT Tech Day highlighted the company's broader portfolio across high-speed interconnects, power and cooling. Lin said its 800G and 1.6T optical transceivers have secured design wins with leading global server brands and are moving into commercial deployment.
FIT is also showcasing 224G interconnects, 800VDC connectors, liquid-cooling UQD products and high-density fiber connectivity for CPO and NPO, targeting rising AI data-center requirements for bandwidth, power density and cooling.
Supply-chain sources said optical communications are increasingly converging with copper cables, high-speed connectors, fiber interfaces, power and liquid cooling at the rack level.
For suppliers, the key shift is from selling individual components to participating earlier in next-generation AI platform design. Companies able to combine signal, power, cooling, and mechanical design into manufacturable system-level solutions will be better positioned as rack architectures evolve.
Article
translated by Levi Li and edited by Jerry Chen
AI Search
Related stories
AI data centers push liquid cooling into core infrastructure
Data center power and cooling overhaul will reshape global AI infrastructure
Delta targets next-gen AI data centers with 800V DC power and liquid cooling solutions
Analysis: AI server power shifts to rack-level architectures as Delta, Liteon take lead
OCP Summit: AI data centers embrace high-voltage power and liquid cooling
Show more
Categories
IT + CE Server, IPC, cloud computing, IoT
Tags
cooling CPO data design FIT switch
Share this article
Other links
Related topic
Global supply chain: Mobile, telecom, computing
BIZ FOCUS
Sep 14, 08:00
Hauman Unveils Micro LED Testing Breakthrough for AI Optical Interconnects
Monday 14 September 2026
Scaling Photonics: Assembly, Test and Fiber Automation at SEMICON Taiwan
Friday 11 September 2026
DAS Expands in Taiwan and Unveils World First at SEMICON
Friday 11 September 2026
Toward Technologies Redefines Relay Solutions at SEMICON Taiwan 2026
MOST-READ
7 DAYS NEWS
Google's v10 TPU opens door to AMD without sidelining Broadcom, MediaTek
Update: ABF substrate maker Unimicron searched in origin-labeling probe
Qualcomm's HBC play pulls Samsung and SK hynix into a new race beyond HBM
SK keyfoundry expands 8-inch capacity as China demand strains supply
Samsung, SK Hynix step up China NAND upgrades as US tool risks grow
Column: When digital trust breaks — quantum risk rewrites supply chain resilience
Wuliangye skips carmaking, turns liquor waste into silicon-carbon anodes
The 1,000-watt problem: a TSMC veteran on the brute force forging Nvidia's AI chips
eCloudvalley expands enterprise AI Agent portfolio
CXMT rides AI boom to 80% operating margin
Full list
Members only
×
Sorry, the page you are trying to open is available only for our paid subscribers.
Please login to read more
New users, please register first
Notification
×
决策记录链 · 只追加
每次领取产生一条记录,重新决策追加一条并以 supersedes 指向被替代的记录(ADR 0014)。
这条决策早于决策账本(2026-10-08 起逐条追加),没有追加记录。