谷歌被曝研发Frozen v2专用AI服务器芯片，2028年前后提升Gemini推理能效但量产与亚洲供应链未定

# 美股盘后催化 5/11: Google研发Frozen v2专用AI服务器芯片，目标Gemini token/瓦效率较现有芯片提升6-10倍

- 美股会话: 2026-07-20
- 方向: MIXED；类型: 产品/技术；规划重大性: 8/10
- 映射模式: competitive
- 美股驱动（规划器初判，需独立核验）: 事实：Reuters、CNBC、The Information和TechCrunch均报道Google开发代号Frozen v2的新AI服务器芯片，目标将Gemini运行效率较现有自研AI芯片提升6-10倍，最早约2028年部署；Alphabet当日上涨。推断：利好Google AI成本控制和自研ASIC叙事，也可能给台积电/ASIC设计/先进封装带来增量；但中长期对外部通用GPU需求和部分AI硬件供应链构成替代风险，且项目仍实验性、量产规模未知。

## 本催化相关美股真实收盘行情

- 英伟达 (NVDA.US): 203.28 +0.23%
- 博通 (AVGO.US): 378.16 +1.98%
- 台积电ADR (TSM.US): 402.3 +0.99%
- 谷歌 (GOOGL.US): 351.99 +1.51%
- Marvell (MRVL.US): 194.94 +3.32%

## 亚太映射假设（仅候选，必须查证）

- 台积电 (2330.TW) [TW/供应商/positive]: 若Frozen v2量产，先进制程和先进封装需求可能增加。；核验问题: Frozen v2代工厂、制程节点、封装方式和量产时间。
- 世芯-KY (3661.TW) [TW/供应商/mixed]: hyperscaler自研ASIC趋势可能利好ASIC设计服务/IP生态，但Google项目参与方未知。；核验问题: Frozen v2是否外包设计服务，台系ASIC服务商或IP公司是否参与。
- 三星电子 (005930.KS) [KR/供应商/竞争替代/mixed]: 若参与代工、HBM或封测则受益；若订单由台积电主导且替代GPU，则间接影响不确定。；核验问题: Frozen v2晶圆代工、内存类型和封装测试供应链。

## 必答问题

- Frozen v2是推理专用还是训练/推理兼用？
- 是否由Broadcom等外部ASIC伙伴参与？
- 2028部署规模是否足以影响Nvidia GPU或现有TPU采购路径？

## 原始证据全文（10 条）

### [E0059] 财联社电报 | 2026-07-20 13:33 UTC | 盘中
财联社7月20日电，谷歌母公司Alphabet上涨2%，此前有报道称谷歌正在研发用于运行人工智能模型的新芯片。
URL: https://www.cls.cn/detail/2431777
关于我们
网站声明
联系方式
用户反馈
网站地图
帮助
首页
电报
话题
盯盘
VIP
FM
投研
下载
全部
加红
公司
看盘
港美股
基金
提醒
2026年07月20日 21:33:56
财联社7月20日电，谷歌母公司Alphabet上涨2%，此前有报道称谷歌正在研发用于运行人工智能模型的新芯片。
阅6179
关联话题
环球市场情报
16.36W 人关注
美股动态
6.43W 人关注
关于我们
|
网站声明
|
联系方式
|
用户反馈
|
网站地图
|
友情链接
|
举报电话：021-54679377转617
举报邮箱：editor@cls.cn
财联社
©2018-2026
上海界面财联社科技股份有限公司 版权所有
沪ICP备14040942号-9
沪公网安备31010402006047号
互联网新闻信息服务许可证：31120170007
沪金信备 [2021] 2号

### [E0213] Reuters Technology | 2026-07-20 13:58 UTC | 盘中
Google plans new chip to run Gemini models more efficiently, the Information reports
URL: https://www.reuters.com/business/google-plans-new-chip-run-gemini-models-more-efficiently-information-reports-2026-07-20/
Google is developing a new server chip that would incorporate elements of its Gemini model directly ​into the hardware, in a bid to serve ‌its AI models more efficiently to users, the Information reported on Monday, citing people familiar with the matter.

The Alphabet-owned (GOOGL.O) company expects the new ​chip, informally dubbed "Frozen v2," to help address an AI ​computing capacity crunch that has fueled internal tensions and ⁠prompted Google Cloud to decline deals with outside customers, the ​report said.

Shares of Alphabet were up 3.3% in early trading.

Here are ​some details:

Google plans to deploy the chip as soon as 2028, though engineers are still finalizing its design and the amount of model information ​that will be hardwired, the report said.

The chip could be ​six to 10 times more efficient than Google's latest custom AI chips ‌based ⁠on the number of AI tokens served per unit of power, according to the report.

"Our teams are constantly researching and experimenting with new innovations... By co-designing our hardware and software from ​the ground up, ​we ensure ⁠our systems are integrated and highly optimized," a Google Cloud spokesperson said.

The 'Frozen' project is aimed ​at creating a new set of homegrown chips ​apart from ⁠Google's tensor processing units (TPUs), rather than replacing them, the report said.

Bloomberg News reporte d last week that Google delayed the launch of its ⁠latest ​Gemini AI model after it fell ​short of internal goals, with the company working to improve its capabilities, particularly ​in coding.

### [E0214] Reuters Technology | 2026-07-20 13:58 UTC | 盘中
Google plans new chip to run Gemini models more efficiently, the Information reports
URL: https://www.reuters.com/business/google-plans-new-chip-run-gemini-models-more-efficiently-information-reports-2026-07-20/
Google is developing a new server chip that would incorporate elements of its Gemini model directly ​into the hardware, in a bid to serve ‌its AI models more efficiently to users, the Information reported on Monday, citing people familiar with the matter.

The Alphabet-owned (GOOGL.O) company expects the new ​chip, informally dubbed "Frozen v2," to help address an AI ​computing capacity crunch that has fueled internal tensions and ⁠prompted Google Cloud to decline deals with outside customers, the ​report said.

Shares of Alphabet were up 3.3% in early trading.

Here are ​some details:

Google plans to deploy the chip as soon as 2028, though engineers are still finalizing its design and the amount of model information ​that will be hardwired, the report said.

The chip could be ​six to 10 times more efficient than Google's latest custom AI chips ‌based ⁠on the number of AI tokens served per unit of power, according to the report.

"Our teams are constantly researching and experimenting with new innovations... By co-designing our hardware and software from ​the ground up, ​we ensure ⁠our systems are integrated and highly optimized," a Google Cloud spokesperson said.

The 'Frozen' project is aimed ​at creating a new set of homegrown chips ​apart from ⁠Google's tensor processing units (TPUs), rather than replacing them, the report said.

Bloomberg News reporte d last week that Google delayed the launch of its ⁠latest ​Gemini AI model after it fell ​short of internal goals, with the company working to improve its capabilities, particularly ​in coding.

### [E0236] CNBC Top News | 2026-07-20 14:00 UTC | 盘中
Alphabet stock pops on report it's developing a more efficient AI chip
URL: https://www.cnbc.com/2026/07/20/alphabet-googl-stock-ai-chip-report.html
The new AI chip, called "Frozen v2," would embed parts of Gemini's architecture directly into the silicon, according to the report.

### [E0237] CNBC Top News | 2026-07-20 14:00 UTC | 盘中
Alphabet stock pops on report it's developing a more efficient AI chip
URL: https://www.cnbc.com/2026/07/20/alphabet-googl-stock-ai-chip-report.html
The new AI chip, called "Frozen v2," would embed parts of Gemini's architecture directly into the silicon, according to the report.

### [E0766] The Information | 2026-07-20 15:56 UTC | 盘中
Alphabet stock jumped 3% on Monday morning after The Information reported that Google was working on a new chip that could be 6 to 10 times more efficient than its current line of homegrown AI chips.

The new chip, internally dubbed “Frozen
URL: https://www.theinformation.com/briefings/alphabet-stock-jumps-information-report-new-google-chip
Alphabet stock jumped 3% on Monday morning after The Information reported that Google was working on a new chip that could be 6 to 10 times more efficient than its current line of homegrown AI chips.

The new chip, internally dubbed “Frozen v2,” works by etching the blueprint for Google’s Gemini AI model into the silicon, making the chip faster and more efficient by cutting down on the number of decisions necessary when running Gemini models. Google expects that the chip could launch as early as 2028.

If successful, the Frozen chip could help address Google’s compute crunch. On recent earnings calls, Google executives have said that its cloud computing business is capacity constrained. But the chip also represents a bet on Google’s current model architecture: While some changes will be possible, including updating model weights, the chip will be usable for future models only if they keep using the same underlying architecture.

### [E1118] 科创板日报 | 2026-07-20 18:06 UTC | 盘中
谷歌被曝正打造新型AI服务器芯片 目标大幅提升Gemini运行效率
URL: https://www.chinastarmarket.cn/detail/2431896
①谷歌研发内部代号“Frozen v2”的服务器芯片，旨在更高效运行Gemini模型，不会取代其通用型TPU; ②消息公布后，谷歌A早盘一度涨近3.7%。谷歌回应称，该项目为试验性探索，暂无大规模量产计划。
财联社7月21日讯（编辑 赵昊）最新消息显示，谷歌正在研发一款内部代号为“Frozen v2”的新型服务器芯片，旨在更高效地运行Gemini模型。
据报道，这款芯片将把Gemini模型部分架构永久嵌入到芯片之中，从而减少回答用户问题所需的计算量和数据传输量。
报道写道，谷歌工程师预计，与公司最新一代AI芯片——TPU（张量处理器）相比，Frozen芯片单位功耗可提供6至10倍的Token处理能力。
报道称，谷歌计划于2028年部署Frozen这款芯片，但不会取代通用型TPU，而是成为谷歌定制芯片产品线中更具针对性的专用分支。
受该消息影响，谷歌A（股票代码：GOOGL）美股早盘一度涨近3.7%，谷歌C（GOOG）最高时涨约3.9%
谷歌在声明中回应称，公司团队“持续研究并试验各类创新技术，旨在为用户和客户提供最佳性能和最高效率”。
声明提到，“虽然并非所有项目都会进入量产阶段，但这种严格的探索是我们全栈战略的核心组成部分。”
谷歌还表示：“通过从底层协同设计硬件和软件，我们确保整个系统实现深度集成，并针对真实工作负载进行了高度优化。”
分析认为，该项目旨在缓解谷歌内部日益严重的算力短缺问题。此前，算力不足不仅加剧了内部资源竞争，据称还迫使谷歌云拒绝部分外部客户业务。
就在上月，谷歌还同意每月向SpaceX支付近10亿美元，以弥补算力缺口，兑现对企业客户的算力承诺。
不过，Frozen的设计也存在一定代价，即灵活性下降。报道称，只有在谷歌未来Gemini模型继续沿用相同底层架构的情况下，该芯片才能兼容后续版本。
目前，谷歌将Frozen v2视为一次试验性项目，暂无计划像TPU一样进行大规模生产。
当下，谷歌AI业务还面临更紧迫的挑战。先前有消息称，下一代Gemini Pro发布时间已经推迟，谷歌还流失了多名资深AI研究人员至竞争对手。

### [E1270] CNBC Top News | 2026-07-20 19:24 UTC | 盘中
2 things capping Monday's market — plus, Alphabet's new AI chip roadmap
URL: https://www.cnbc.com/2026/07/20/2-things-capping-monday-market-rally-alphabet-new-ai-chip-plans.html
Every weekday, the CNBC Investing Club with Jim Cramer releases the Homestretch — an actionable afternoon update, just in time for the last hour of trading on Wall Street. Stocks moved slightly lower on Monday afternoon, as an initial S & P 500 rally failed to hold. While the market debated whether the recent correction in AI stocks had gone too far or if volatility would continue, oil prices and rising bond yields remained headwinds for equities. WTI crude traded above $83 per barrel during the session after briefly dipping to roughly $80 earlier Monday. Treasury yields continued their recent advance, with the 10-year climbing back to around 4.6%. Federal Reserve interest rate-hike probabilities crept higher on the renewed threat of an uptick in inflation due to energy prices going up again. Alphabet shares rose on a report that its Google unit is developing a more efficient AI chip. The Information said Monday that the tech giant is working on a new server chip designed to make its Gemini AI models run more efficiently. The chips, called Frozen v2, aren't expected to be deployed until 2028 and won't replace its tensor processing units (TPUs), which are co-designed by fellow Club name Broadcom . But the Frozen chips could help Alphabet run its AI models more efficiently by generating more tokens per unit of power. Tokens are units of data processed by AI models. While still potentially years away from releasing this new chip, the Information story highlights Google's leadership in developing its own chips, which are both cost-effective and reduce reliance on third-party accelerators like Nvidia . Alphabet's recent loss of high-profile AI talent and the delay of its newest, most powerful AI model have raised some concerns about its AI efforts. But projects like Frozen v2 should help reassure investors that the company continues to invest in its long-term AI strategy to keep its edge. There are no major after-the-bell earnings . A busy slate of quarterly results is on tap before Tuesday's opening bell, with reports from 3M, Danaher, Halliburton, Charles Schwab, General Motors, Novartis, and Northrop Grumman. The first earnings report of the week in the portfolio is Capital One , which is set for Tuesday evening. We see the release from the credit card giant as a chance to show investors that its $35 billion Discover deal was worth it. There are no major economic reports on Tuesday. (See here for a full list of the stocks in Jim Cramer's Charitable Trust, including GOOGL, NVDA, COF.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust's portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.

### [E1579] TechCrunch AI | 2026-07-20 21:21 UTC | 收盘后
Google is working on a new AI chip designed to make Gemini more efficient
URL: https://techcrunch.com/2026/07/20/google-is-working-on-a-new-ai-chip-designed-to-make-gemini-more-efficient/
Alphabet, Google’s parent company, is designing a new server chip to help its in-house Gemini models operate more efficiently.
The new chip, internally dubbed “Frozen v2,” is slated to be released sometime in 2028, The Information reported, citing anonymous sources. According to the report, the chip could be between six and 10 times more efficient than Google’s existing AI chips, measured by the number of tokens generated per unit of power.
In a response to TechCrunch, the company didn’t directly confirm the report. It didn’t deny it either.
“Our teams are constantly researching and experimenting with new innovations to deliver maximum performance and efficiency for our users and customers,” Google told TechCrunch. “While not every project moves into production, this rigorous exploration is central to our full stack approach. By co-designing our hardware and software from the ground up, we ensure our systems are integrated and highly optimized for real-world workloads.”
AI companies have increasingly sought to produce their own chips as a way to make their in-house models run more efficiently and to address global shortages in AI computing capacity. Such efficiency has become a key selling point for tech companies as concerns about AI spend have dampened the market euphoria that previously characterized the industry. At the same time, firms are engaged in an ongoing attempt to wean themselves off chipmaker Nvidia, which has historically dominated the AI chip market and whose dominance has left major AI makers dependent on its hardware.
In June, OpenAI announced its first custom chip, an inference processor dubbed Jalapeño. Earlier this month, it was reported that Anthropic was discussing a new chipmaking partnership with Samsung.
Investors have previously worried about Alphabet’s massive planned expenditures designed to help it build out its AI strategy. Earlier this year, Google said that it plans to spend between $180 billion and $190 billion. With so much money at stake, the company needs to prove that those investments will pay off.
News of the more efficient Frozen v2 chip appears to have assuaged investors, giving Google a boost ahead of its earnings report later this week. Following publication of The Information’s report, the company’s stock climbed some 3% on Monday morning.

### [E1580] TechCrunch AI | 2026-07-20 21:21 UTC | 收盘后
Google is working on a new AI chip designed to make Gemini more efficient
URL: https://techcrunch.com/2026/07/20/google-is-working-on-a-new-ai-chip-designed-to-make-gemini-more-efficient/
Alphabet, Google’s parent company, is designing a new server chip to help its in-house Gemini models operate more efficiently.
The new chip, internally dubbed “Frozen v2,” is slated to be released sometime in 2028, The Information reported, citing anonymous sources. According to the report, the chip could be between six and 10 times more efficient than Google’s existing AI chips, measured by the number of tokens generated per unit of power.
In a response to TechCrunch, the company didn’t directly confirm the report. It didn’t deny it either.
“Our teams are constantly researching and experimenting with new innovations to deliver maximum performance and efficiency for our users and customers,” Google told TechCrunch. “While not every project moves into production, this rigorous exploration is central to our full stack approach. By co-designing our hardware and software from the ground up, we ensure our systems are integrated and highly optimized for real-world workloads.”
AI companies have increasingly sought to produce their own chips as a way to make their in-house models run more efficiently and to address global shortages in AI computing capacity. Such efficiency has become a key selling point for tech companies as concerns about AI spend have dampened the market euphoria that previously characterized the industry. At the same time, firms are engaged in an ongoing attempt to wean themselves off chipmaker Nvidia, which has historically dominated the AI chip market and whose dominance has left major AI makers dependent on its hardware.
In June, OpenAI announced its first custom chip, an inference processor dubbed Jalapeño. Earlier this month, it was reported that Anthropic was discussing a new chipmaking partnership with Samsung.
Investors have previously worried about Alphabet’s massive planned expenditures designed to help it build out its AI strategy. Earlier this year, Google said that it plans to spend between $180 billion and $190 billion. With so much money at stake, the company needs to prove that those investments will pay off.
News of the more efficient Frozen v2 chip appears to have assuaged investors, giving Google a boost ahead of its earnings report later this week. Following publication of The Information’s report, the company’s stock climbed some 3% on Monday morning.

## 研究交付要求

这是一条已经从整场材料语义拆出的独立催化，不要重新把它压回大盘总评。核实美股异动原因和传播时序；用结构化供应链、内部 KB 与公司披露坐实每个亚洲候选；逐腿查真实行情和本地市场 price-in；候选不成立也要逐条给出有证据的 AVOID。