Investing.comQwen · 实盘不做把握 4/10不做原因:已被定价
北京时间·🕒 发布 05:12:32·⟳ 抓取 05:13:10 +38s·⚡ 决策 08-27 10:55:41 +5h42m·61.3 天前
AWS与英伟达宣布2027-2028年额外部署200万颗GPU，涵盖Blackwell Ultra、Rubin及Rubin Ultra
NVDA盘后已+5.3%完成重定价，200万颗GPU部署在3月GTC已公布的$1T可见度框架内，非增量指引；A股供应链今日已集体上涨2-4%且近20日极端波动，我方历史胜率低（中际旭创35%），部署为2027-2028年远期需求确认而非即时订单，追入风险收益比不佳
📄 原文Investing.com↗ 原文
AWS and Nvidia plan to deploy 2 million additional GPUs by 2028
Nvidia Q2 revenue more than doubles, CEO Huang says AI at its ’inflection point’ SEATTLE and SANTA CLARA, Calif. - Amaz…
不做 英伟达（NVDA.US）
市值$5.07万亿，盘后已+5.3%完成重定价；200万颗GPU部署（约$60-80B/2年）在3月GTC已公布的$1T可见度内，非增量指引变化，超大市值软催化无新金额增量
不做 中际旭创（300308.SZ）
今日已+2.57%反应本催化，近20日极端波动（日涨跌5-13%），我方历史胜率仅35%（20笔），部署为2027-2028年远期需求确认而非即时订单，追入风险收益比不佳
不做 英维克（002837.SZ）
8/25+10%、8/26+3%、8/27+4.3%已连续三日大涨，本次事件为需求确认而非新增订单，追入面临高概率回调
不做 沪电股份（002463.SZ）
今日+3.27%已反应，近20日波动极端（-8%至+10%），PCB为通用算力基建非本次事件特异性受益，无独立催化
不做 新易盛（300502.SZ）
今日+2.87%已反应，近20日极端波动（-12%至+14%），光模块需求确认为远期非即时，无独立增量信息
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
AWS宣布2027-2028年额外部署200万颗Nvidia GPU，方向明确利好算力产业链。但按$30-40K/颗估算$60-80B，在黄仁勋3月GTC已公布的'2025-2027年≥$1T Blackwell+Rubin收入'可见度内，非增量惊喜。NVDA盘后已涨5.3%完成定价；A股光模块/液冷/PCB今日已涨2-4%且近20日日波动5-13%，追入面临极高回撤风险。部署时间线为2027-2028年，非即时收入确认。综合判断无正期望交易窗口。
🔎 新鲜度:最早出现 2026-08-26 · 扩散 medium · price-in 风险 high ↗
消息传播链
- Investing.com为聚合站转载，原文应为Nvidia/AWS官方新闻稿；因Web搜索不可用未能定位原始首发链接市场反应:NVDA盘后+4.7%→+5.3%；SKHY ADR盘后+5%
- 2026-08-27市场反应A股市场非一手A股算力供应链开盘即涨：中际旭创+2.6%、新易盛+2.9%、英维克+4.3%、沪电+3.3%市场反应:板块整体上涨2-4%
关键理由
- AWS 200万颗GPU部署确认2027-2028年AI算力需求持续强劲，利好GPU全产业链
- 部署含Rubin/Rubin Ultra新品，验证下一代产品需求可见度
- 为美国政府建10万颗GPU安全AI工厂，打开政府市场增量
主要风险
- 200万颗部署在既有$1T可见度内，市场可能视为确认而非惊喜，后续缺乏上修空间
- A股供应链近20日极端波动（日涨跌5-13%），追入后面临高概率大幅回撤
反向证据
- 做多前查反向 · 来源 query_modeling_kb · 日期 2026-03-18 · 影响 weakens
黄仁勋GTC2026已公开'through 2027 at least $1T'可见度，Evercore确认21个月后还会上修；本次200万颗（$60-80B）在既有框架内
检索:NVDA AWS 2 million GPU already expected consensus estimate - 做多前查反向 · 来源 search_case_library · 影响 weakens
我方在中际旭创上20笔已结算交易胜率仅35%，中位收益-1.07%，追高模式历史上亏损为主
检索:中际旭创 历史胜率 追高回撤
先例如何影响了这次决策
对本次决策的影响:先例缺失不影响本次AVOID结论，因为主标的已盘后重定价、A股已开盘反应且波动极端，无论先例如何都无法建立正期望入场
方向:无法完成先例研究
涨幅量级:无
形态:无
工具调用 · 历史记录,结果未知(8 种)
- fetch_decision_experience · 结果未知
- search_processed_events · 结果未知
- web_search · 结果未知
- stock_quote · 结果未知
- query_modeling_kb · 结果未知
- supply_chain_lookup · 结果未知
- kb_search · 结果未知
- search_case_library · 结果未知
原文全文
展开原文 · 2,156 字
Nvidia Q2 revenue more than doubles, CEO Huang says AI at its ’inflection point’
SEATTLE and SANTA CLARA, Calif. - Amazon Web Services and Nvidia announced today an expansion of their partnership to deploy 2 million additional Nvidia GPUs across AWS’s global infrastructure in 2027-2028, according to a press release statement.
The deployment will include Nvidia Blackwell Ultra, Rubin and Rubin Ultra GPUs. AWS (NASDAQ:AMZN) will also expand Nvidia (NASDAQ:NVDA) Blackwell capacity, including RTX PRO 4500 Blackwell Server Edition GPUs for Amazon EC2 G7 instances. AWS is the first major cloud provider to offer compute instances accelerated by RTX PRO 4500.
The companies plan to build AI factories for the U.S. government, including 100,000 GPUs on secure AWS infrastructure for federal and national-security workloads classified at Impact Level 6 and above.
AWS and Nvidia are working to bring Vera CPU-based infrastructure to AWS to support AI workloads requiring high-performance CPU compute alongside accelerated infrastructure. The companies are extending Nvidia NVLink Fusion with custom Nvidia high-bandwidth memory technology, which would give AWS’s Trainium chips access to faster, more power-efficient memory.
"Customers want the freedom to choose the best tools for their AI workloads," said Matt Garman, CEO of AWS.
Jensen Huang, founder and CEO of Nvidia, said the companies are expanding their partnership "across the full stack - GPUs, CPUs, networking, open models and software."
The collaboration includes GPU-accelerated data processing on Amazon EMR using Amazon EC2 G7 instances, delivering up to 3.7 times faster processing speeds and 30% better price performance compared to CPU-based configurations. GPU-accelerated vector indexing on Amazon OpenSearch Service delivers up to nine times faster vector indexing at a quarter of the cost.
Amazon Robotics is integrating Nvidia’s physical AI platform for next-generation robot development. Nvidia’s Nemotron open models are available on Amazon Bedrock and Amazon SageMaker.
This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.
SEATTLE and SANTA CLARA, Calif. - Amazon Web Services and Nvidia announced today an expansion of their partnership to deploy 2 million additional Nvidia GPUs across AWS’s global infrastructure in 2027-2028, according to a press release statement.
The deployment will include Nvidia Blackwell Ultra, Rubin and Rubin Ultra GPUs. AWS (NASDAQ:AMZN) will also expand Nvidia (NASDAQ:NVDA) Blackwell capacity, including RTX PRO 4500 Blackwell Server Edition GPUs for Amazon EC2 G7 instances. AWS is the first major cloud provider to offer compute instances accelerated by RTX PRO 4500.
The companies plan to build AI factories for the U.S. government, including 100,000 GPUs on secure AWS infrastructure for federal and national-security workloads classified at Impact Level 6 and above.
AWS and Nvidia are working to bring Vera CPU-based infrastructure to AWS to support AI workloads requiring high-performance CPU compute alongside accelerated infrastructure. The companies are extending Nvidia NVLink Fusion with custom Nvidia high-bandwidth memory technology, which would give AWS’s Trainium chips access to faster, more power-efficient memory.
"Customers want the freedom to choose the best tools for their AI workloads," said Matt Garman, CEO of AWS.
Jensen Huang, founder and CEO of Nvidia, said the companies are expanding their partnership "across the full stack - GPUs, CPUs, networking, open models and software."
The collaboration includes GPU-accelerated data processing on Amazon EMR using Amazon EC2 G7 instances, delivering up to 3.7 times faster processing speeds and 30% better price performance compared to CPU-based configurations. GPU-accelerated vector indexing on Amazon OpenSearch Service delivers up to nine times faster vector indexing at a quarter of the cost.
Amazon Robotics is integrating Nvidia’s physical AI platform for next-generation robot development. Nvidia’s Nemotron open models are available on Amazon Bedrock and Amazon SageMaker.
This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.
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。