AMD 评级格局(截至 5/8 收盘)
数据来源:MarketBeat 44 位分析师汇总
本期合并覆盖 2026-05-09 至 05-11(约 45 小时),横跨周末及周一早间。三条核心主线并进:①腾讯云 5/9 正式涨价 5% 落地,阿里云/百度云/AWS/Google Cloud 全线涨价,国金证券判断「2026 年算力全链通胀」成立;②AMD Q1 财报后 17 家机构联动——5 家升级(GS Buy $450、Bernstein Outperform $525)、4 家谨慎(Jefferies Hold/Oppenheimer/Raymond James/Wolfe),PE 137x 已超共识目标价 17%,Bull-Bear 结构性分化持续;③高盛「Tracking Trillions」预测 AI CapEx $7.6T 基线,同时 Covello 判断「FOMO 比 ROI 更顽固」;摩根士丹利将 2026 年 Hyperscaler CapEx 上调至 $805B,AI 基建融资进入结构性通道。国内方面,广发/国泰海通/中信三大券商发布研报,方向一致指向国产算力适配加快与光通信景气上行。
本期覆盖时间:2026-05-09 02:38 → 2026-05-11 00:00(北京时间),约 45 小时窗口,横跨周六周日及周一早间。因上期(5/10)系统调度失败未能出刊,本期合并补覆盖,内容密度高于常规日报。 市场状态:A 股、港股 5/9-5/10 休市;美股周一(5/11 北京时间对应 5/10 美东)正常交易。
"We upgrade AMD to Buy with a 12-month price target of $450 given tailwinds to its server CPU business we expect from agentic AI, as well as upside in datacenter GPUs in 2027 and beyond."——「我们将 AMD 升至买入,12 个月目标价 $450,因其服务器 CPU 业务预期受益于 agentic AI 顺风,以及 2027 年及以后数据中心 GPU 的上行空间。」
"We have continually been surprised not only by the strength of AMD's business but also by the strength of the stock. That being said, while many stocks have been climbing strictly on vibes lately the company deserves significant credit for a fundamental story that increasingly is looking real."——「我们不断被 AMD 业务的强劲以及股价的表现所震惊。话虽如此,近期许多股票纯靠情绪上涨,但 AMD 确实值得充分认可——其基本面故事越来越像真的。」
"We've struggled somewhat to independently corroborate management's 2H outlook; however, we see strength in CPUs as offering a bit of insurance if GPUs ramp somewhat behind current expectations."——「我们难以独立核实管理层的下半年 GPU 展望,但 CPU 强势可以作为一种保险——若 GPU 爬坡略低于预期,可通过重新分配产能对冲。」
"While we are constructive on AMD's CPU opportunity on agentic AI demand, we await the release and success of its next gen GPU product MI450 and Helios racks."——「对于 AMD CPU 受益 agentic AI 需求的机会,我们持建设性态度,但需等待 MI450 下一代 GPU 产品及 Helios 机架的发布和成功验证。」
「国内算力黄金年代开启,供给紧缺下好用的算力本身就是最核心的资产。」
"Why be positive when you can be cynical? This seems like a major heat check before the IPO."——「能泼冷水为何要说好话?这看起来像是 IPO 前的一次重大热度检查。」
"FOMO has proven a stronger incentive than poor stock performance as hyperscalers have prioritized being involved in the AI arms race over their current shareholders."——「事实证明,FOMO 是比糟糕的股票表现更强的驱动力——hyperscaler 优先参与 AI 军备竞赛,而非服务当前股东。」
"Something has to change with this dynamic — either the companies higher in the chain need to start earning a return on investment or they will eventually need to spend less on the chips that are powering this build."——「这种动态必须改变——要么产业链上游公司开始获得投资回报,要么最终不得不减少对驱动这场建设的芯片支出。」
说明:本期窗口覆盖周末,国内头部券商研报密度低于工作日。采集到广发证券 3 篇、国泰海通 2 篇、中信证券 1 篇,均为 5/10-5/11 发布。腾讯云涨价专项点评截至本期截稿时仍未出现,预计 5/11(周一)开盘后陆续跟进。
"Google is one of the two best-positioned AI companies because they own most of the stack — chips, models, infrastructure and distribution. On top of that, they're nicely profitable."——「Google 是 AI 领域位置最好的两家公司之一,因为他们拥有全栈——芯片、模型、基础设施和分发渠道。此外,他们还在盈利。」
"There are so many great, amazing foundation model companies, and we try to invest in all of them. We don't pick winners. We need to support everyone."——「有太多出色的基础模型公司,我们尽量投资所有公司。我们不挑选赢家,需要支持每一家。」
"The old 'China is six to nine months behind' frame for agentic coding is no longer defensible. The remaining gap is narrow, contested, and now decided by the evaluator, the scaffold, and the benchmark, not by raw capability."——「旧的'中国落后 6-9 个月'框架对 agentic coding 不再成立。剩余差距是狭窄的、有争议的,现在由评估者、scaffold 和基准决定,而非原始能力。」
"I think this year will be the year of agents... The smartphone is not going to go anywhere, but certain workloads are going to be more intuitive to you to just do it [on other devices]."——「我认为今年将是 Agent 之年……智能手机不会消失,但某些工作负载在其他设备上会更自然。」
このコンテンツについて、さらに観点や背景を補足しましょう。