📰 DigiTimesDigiTimesDoubao · 实盘不做把握 7/10不做原因:无增量信息
北京时间·🕒 发布 07:38:34·⟳ 抓取 07:50:06 +11m32s·⚡ 决策 09-15 13:11:10 +5h21m·42.2 天前
DigiTimes 回顾英伟达35亿美元投资联发科可转债及双方20年芯片合作史，无新增交易事实
这是一篇产业媒体的合作史回顾稿，核心催化（英伟达35亿美元认购联发科可转债并扩展AI芯片合作）发生在2026年8月底、已约3周，市场早已消化；全文为历史叙事与并购史梳理，未含任何新订单、新金额、新指引或官方新确认，缺乏可交易的增量事实与时间edge。
📄 原文DigiTimes↗ 原文
Nvidia, MediaTek deepen 20-year chip ties
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 联发科（2454.TW）
35亿美元英伟达可转债投资为8月底旧闻、已被定价；本篇DigiTimes为3周后的历史回顾稿，无任何新增交易事实或时间edge，不构成新的做多催化。
不做 英伟达（NVDA.US）
投资金额对英伟达体量影响微小且为旧闻，本文仅回顾并购史、无新增信息，无可交易edge。
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此文重述英伟达8月底35亿美元投资联发科可转债的旧闻并回顾双方20年芯片合作史，没有任何新增交易事实或数据，催化已在3周前释放，不具备短线edge，故回避。
🔎 新鲜度:最早出现 2026-08-31 · 扩散 high · price-in 风险 high ↗
关键理由
- 核心利好（35亿美元可转债+定制AI芯片合作）为2026年8月底既有事件，本篇仅历史回顾，无新增信息可驱动新一轮重定价。
主要风险
- 若后续出现该合作的具体订单/量产/分成等新公告，应作为独立新事件重新研判，不能据此回顾稿提前布局。
先例如何影响了这次决策
对本次决策的影响:支持以NO_INCREMENTAL_FACT回避，不产生任何交易腿。
方向:无可对照的新增事件；本篇为旧闻回顾，不需要以历史先例校准涨幅。
涨幅量级:无新增催化，不设定止盈幅度。
形态:产业媒体在旧交易数周后发布的历史回顾通常不带来新的同向重定价，更多是背景传播。
工具执行结果 · 成功 1 · 降级 0 · 失败 0 · 取消 0
- search_processed_events 成功 stock_tools HTTP · 49 ms
{"result": "search_processed_events:agent query=`nvidia mediatek 3.5 billion convertible bonds investment | nvidia mediatek custom ai chip automotive data center partnership august 2026 | 英伟达 联发科 35亿美元 可转债 投资 合作`；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 表达式没有召回候选，
原文全文
展开原文 · 7,999 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
Semiconductors
Nvidia, MediaTek deepen 20-year chip ties
Monica Chen
, Taipei
Sep 15, 2026, 07:38
0
Credit: DIGITIMES
Nvidia and MediaTek have recently drawn market attention after Nvidia invested US$3.5 billion in MediaTek convertible bonds, expanding cooperation into custom AI chips, AI PCs, automotive, and data centers. But their overlap began more than 20 years ago, during competition in the PC chipset market, when MediaTek, ALi, ULi, and Nvidia were already connected.
ALi, which was originally part of Acer before spinning off as an independent company in 1996, split off its PC chipset division in 2002 to form ULi. In 2004, control of ALi changed hands, MediaTek became a major shareholder, and ULi was indirectly brought into MediaTek's orbit. At the end of 2005, Nvidia agreed to acquire ULi for about US$52 million, gaining southbridge and chipset design capabilities.
At the time, ALi held 36.8% of ULi and UMC held 28%, while MediaTek held 25% of ALi. After Nvidia's acquisition of ULi, MediaTek was also indirectly tied to the deal.
ULi was founded amid fierce competition in the PC chipset market, where Intel, AMD, ATI, Nvidia, VIA Technologies, and SiS all had a presence. ULi entered the market with northbridge and southbridge chipsets and, lacking graphics core technology, partnered with ATI to move into OEM and channel markets. But with its product line and scale still limited, ULi proactively approached Nvidia CEO Jensen Huang. Nvidia then agreed to acquire ULi in December 2005 for a total of about US$52 million.
At the time, Nvidia had already built an advantage in the GPU market and was aggressively developing chipset and platform products. Platform efforts such as SLI also required chipset support. Huang said Nvidia needed to keep strengthening its chipset design capabilities, and that ULi's team would help make its platform strategy more complete. For ULi, the deal was a lifeline, and it also became a notable case of a Taiwanese chipset maker being acquired by an international giant that year.
After Nvidia acquired ULi, ALi was expected to book profit of about NT$230 million (approx. US$7.2 million), while MediaTek also benefited indirectly through its stake in ALi. However, MediaTek and Nvidia had not yet formed the direct partnership seen today.
Failed M&A talks a decade ago
More than 10 years ago, the two sides reportedly explored an acquisition, but the talks did not succeed. Nvidia had by then become the market leader in standalone PC graphics cards and was investing in gaming, virtual reality, deep learning, and autonomous driving, with fiscal 2016 revenue of about US$5.01 billion. The deal reportedly fell through because Nvidia was continuing to invest in CUDA and AI technologies, and even though mobile device growth had not met expectations, Huang had no intention of selling.
In recent years, Nvidia GPUs have grown rapidly in the AI computing market, while CUDA has gradually built out a complete software ecosystem. MediaTek, meanwhile, has expanded beyond smartphone SoCs into custom chips for automotive, AI PCs, and data centers, and the two companies' cooperation has gradually extended from product collaboration to AI computing platforms.
Nvidia's US$3.5 billion investment in late August 2026 deepened its relationship with MediaTek. Compared with the US$52 million ULi acquisition in 2005, the two deals differ by about 67x in value.
The bigger difference is strategic: in 2005, Nvidia acquired southbridge, chipset technology, and an R&D team. In 2026, as AI chip and architecture competition intensifies, MediaTek is participating with ASIC, SoC, and systems integration capabilities, while Nvidia is using capital investment to build a deeper partnership.
Nvidia broadens its product bets
Looking at Nvidia's acquisitions and investments over the past decade also shows how its product strategy has evolved. The company has moved beyond GPUs into CPUs, networking, software, ASICs, and rack-scale AI systems.
In 2013, it acquired HPC compiler maker PGI to strengthen general-purpose GPU computing and the CUDA development environment. In 2019, it agreed to acquire Mellanox for about US$6.9 billion, gaining InfiniBand, Ethernet, networking chips, and high-speed interconnect capabilities. That move extended Nvidia from GPU accelerators to large-scale GPU clusters, and its DGX, HGX and SuperPOD systems expanded accordingly.
Nvidia announced its roughly US$40 billion bid for Arm in 2020, but the deal collapsed in 2022. CPU capability remained a key gap in its data center strategy. It then moved to develop its own Grace CPU, followed by Grace Hopper and Grace Blackwell products. Its 2021 acquisition of DeepMap also extended Nvidia's AI computing into autonomous driving and automotive markets.
After generative AI surged at the end of 2022, Nvidia moved further into AI models, startups and GPU cloud services, investing in OpenAI and CoreWeave to expand the link between GPU supply and AI computing demand. In 2024, it acquired GPU management software company Run:ai, strengthening resource management and utilization efficiency for large GPU clusters.
As AI inference became a focus in 2025 and 2026, Groq—known for low-latency, high-throughput AI inference architectures that differ from traditional large-scale GPU training markets—became one of Nvidia's targets. Its recent US$12.9 billion agreement to acquire Hugging Face would be its second-largest announced deal, behind the reported US$20 billion Groq asset-and-technology transaction.
Article
translated by Charlene Chen and edited by Ysi Chen
AI Search
Related stories
Nvidia moves to build US$10B AI empire, one acquisition at a time
MediaTek shrugs off Nvidia-Intel deal
Acer targets AI inference market with new launches partnered with MediaTek and Nvidia
Categories
Chips + components IC manufacturing ICT manufacturing IT + CE
Tags
2026 acquisition ALi automotive chips chipset competition GPU investment market MediaTek Nvidia PC southbridge
Companies
ALi MediaTek
Share this article
Other links
Print
Related topics
Global supply chain: CE, IPC, automotive
Global supply chain: China market
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
Semiconductors
Nvidia, MediaTek deepen 20-year chip ties
Monica Chen
, Taipei
Sep 15, 2026, 07:38
0
Credit: DIGITIMES
Nvidia and MediaTek have recently drawn market attention after Nvidia invested US$3.5 billion in MediaTek convertible bonds, expanding cooperation into custom AI chips, AI PCs, automotive, and data centers. But their overlap began more than 20 years ago, during competition in the PC chipset market, when MediaTek, ALi, ULi, and Nvidia were already connected.
ALi, which was originally part of Acer before spinning off as an independent company in 1996, split off its PC chipset division in 2002 to form ULi. In 2004, control of ALi changed hands, MediaTek became a major shareholder, and ULi was indirectly brought into MediaTek's orbit. At the end of 2005, Nvidia agreed to acquire ULi for about US$52 million, gaining southbridge and chipset design capabilities.
At the time, ALi held 36.8% of ULi and UMC held 28%, while MediaTek held 25% of ALi. After Nvidia's acquisition of ULi, MediaTek was also indirectly tied to the deal.
ULi was founded amid fierce competition in the PC chipset market, where Intel, AMD, ATI, Nvidia, VIA Technologies, and SiS all had a presence. ULi entered the market with northbridge and southbridge chipsets and, lacking graphics core technology, partnered with ATI to move into OEM and channel markets. But with its product line and scale still limited, ULi proactively approached Nvidia CEO Jensen Huang. Nvidia then agreed to acquire ULi in December 2005 for a total of about US$52 million.
At the time, Nvidia had already built an advantage in the GPU market and was aggressively developing chipset and platform products. Platform efforts such as SLI also required chipset support. Huang said Nvidia needed to keep strengthening its chipset design capabilities, and that ULi's team would help make its platform strategy more complete. For ULi, the deal was a lifeline, and it also became a notable case of a Taiwanese chipset maker being acquired by an international giant that year.
After Nvidia acquired ULi, ALi was expected to book profit of about NT$230 million (approx. US$7.2 million), while MediaTek also benefited indirectly through its stake in ALi. However, MediaTek and Nvidia had not yet formed the direct partnership seen today.
Failed M&A talks a decade ago
More than 10 years ago, the two sides reportedly explored an acquisition, but the talks did not succeed. Nvidia had by then become the market leader in standalone PC graphics cards and was investing in gaming, virtual reality, deep learning, and autonomous driving, with fiscal 2016 revenue of about US$5.01 billion. The deal reportedly fell through because Nvidia was continuing to invest in CUDA and AI technologies, and even though mobile device growth had not met expectations, Huang had no intention of selling.
In recent years, Nvidia GPUs have grown rapidly in the AI computing market, while CUDA has gradually built out a complete software ecosystem. MediaTek, meanwhile, has expanded beyond smartphone SoCs into custom chips for automotive, AI PCs, and data centers, and the two companies' cooperation has gradually extended from product collaboration to AI computing platforms.
Nvidia's US$3.5 billion investment in late August 2026 deepened its relationship with MediaTek. Compared with the US$52 million ULi acquisition in 2005, the two deals differ by about 67x in value.
The bigger difference is strategic: in 2005, Nvidia acquired southbridge, chipset technology, and an R&D team. In 2026, as AI chip and architecture competition intensifies, MediaTek is participating with ASIC, SoC, and systems integration capabilities, while Nvidia is using capital investment to build a deeper partnership.
Nvidia broadens its product bets
Looking at Nvidia's acquisitions and investments over the past decade also shows how its product strategy has evolved. The company has moved beyond GPUs into CPUs, networking, software, ASICs, and rack-scale AI systems.
In 2013, it acquired HPC compiler maker PGI to strengthen general-purpose GPU computing and the CUDA development environment. In 2019, it agreed to acquire Mellanox for about US$6.9 billion, gaining InfiniBand, Ethernet, networking chips, and high-speed interconnect capabilities. That move extended Nvidia from GPU accelerators to large-scale GPU clusters, and its DGX, HGX and SuperPOD systems expanded accordingly.
Nvidia announced its roughly US$40 billion bid for Arm in 2020, but the deal collapsed in 2022. CPU capability remained a key gap in its data center strategy. It then moved to develop its own Grace CPU, followed by Grace Hopper and Grace Blackwell products. Its 2021 acquisition of DeepMap also extended Nvidia's AI computing into autonomous driving and automotive markets.
After generative AI surged at the end of 2022, Nvidia moved further into AI models, startups and GPU cloud services, investing in OpenAI and CoreWeave to expand the link between GPU supply and AI computing demand. In 2024, it acquired GPU management software company Run:ai, strengthening resource management and utilization efficiency for large GPU clusters.
As AI inference became a focus in 2025 and 2026, Groq—known for low-latency, high-throughput AI inference architectures that differ from traditional large-scale GPU training markets—became one of Nvidia's targets. Its recent US$12.9 billion agreement to acquire Hugging Face would be its second-largest announced deal, behind the reported US$20 billion Groq asset-and-technology transaction.
Article
translated by Charlene Chen and edited by Ysi Chen
AI Search
Related stories
Nvidia moves to build US$10B AI empire, one acquisition at a time
MediaTek shrugs off Nvidia-Intel deal
Acer targets AI inference market with new launches partnered with MediaTek and Nvidia
Categories
Chips + components IC manufacturing ICT manufacturing IT + CE
Tags
2026 acquisition ALi automotive chips chipset competition GPU investment market MediaTek Nvidia PC southbridge
Companies
ALi MediaTek
Share this article
Other links
Related topics
Global supply chain: CE, IPC, automotive
Global supply chain: China market
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。