MOCHA retweetledi
MOCHA
121 posts
MOCHA
@onlyforchengren
AMD 2026 TARGET 320~350 AMD TO THE MOOOOON
Katılım Kasım 2021
67 Takip Edilen8 Takipçiler
MOCHA retweetledi
MOCHA retweetledi
Walking out of his last news conference as Fed Chairman Jay Powell says. “Thank you very much, everyone. I won't see you next time.” #FederalReserve #Powell
English
MOCHA retweetledi
微软云毛利64%，有所下滑。
财报电话会上CFO点名说是因为GitHub Copilot被人薅秃了。
还记得轮子哥 @geniusvczh 推荐的一个请求干一个小时，39套餐一个月用掉1000美元额度。
The Transcript@TheTranscript_
$MSFT CFO: "Microsoft cloud gross margin %ge should be roughly 64%, down YoY driven by continued investments in AI and increased GitHub, co-pilot usage. Just this week, we announced a business model transition in GitHub co-pilot that will align pricing with usage and value."
中文
MOCHA retweetledi
看了很多博主对谷歌的分析，没到点子上，这个推文才是谷歌财报后胜出的真正原因，TPU 8i 在 Agent Workflow 上具有了明显优势，而英伟达 GPU 目前在需要频繁切换上下文的 Agent 任务上只有 30-40%，因为 GPU 诞生之初就是为了高吞吐量适合干类似训练这样的重活。
Agent Workflow 的特性有循环、分支、工具调用、状态管理、多代理协作，所以需要大量 CPU 辅助 GPU，提高利用率，尤其是在数据调度、orchestration 和 pre/post-processing 环节。
不过我觉得还是要看下半年，Rubin 和 TPU 8i 都能上手测评的时候，看看 head-to-head agent 的基准测试才能下定论。
Y Combinator@ycombinator
Inference Chips for Agent Workflows @sdianahu Most AI chips are designed for "prompt in, response out." Agents don't work that way. They loop, branch, and hold context across dozens of steps, and current GPUs hit 30–40% utilization as a result. That gap is where purpose-built silicon wins.
中文
MOCHA retweetledi
Citing UBS's AMD earnings preview:
- The market is clearly aware that INTC's guidance reads very favorably for AMD.
- This is particularly true for server CPU, given commentary implying that INTC is undershipping the market by roughly 20%.
- The key question here is AMD's supply, but our field work throughout the quarter indicated that supply of INTC parts has been far more constrained than that of AMD parts.
- We therefore see a very favorable setup for AMD and expect revenue to be guided up at least $1B Q/Q to the low $11B range (vs. Street consensus of ~$10.4B).
- Intel's CY2026 outlook implies its Data Center & AI segment growing ~40% Y/Y, and we now see AMD server CPU growing as much as 80% this year, with units up ~40–45% and pricing up around 20%.
- From a competitive standpoint, our checks remain constructive. AMD's CPU portfolio continues to compare favorably with Intel's offerings, and the lack of meaningful timeline updates for Diamond Rapids and Coral Rapids reinforces our view that AMD should maintain a competitive advantage across the x86 ecosystem through C2026.
$INTC $AMD
English
MOCHA retweetledi
it is better to be true, i spent over 1mil to roll all my short leg calls.
Scrooge McDuck@ScroogeCap
SemiAnalysis calls out industry chatter that Anthropic is planning to incorporate $AMD GPUs into their hardware stack for the first time starting from MI450. Would expect a lot larger relative outperformance in the coming days if true.
English
MOCHA retweetledi
Someone is adding MILLIONS worth of $SPY & $GLD calls now...
English
MOCHA retweetledi
人的直觉真的很难相信指数增长。
Anthropic 在 2025 年提交给美国政府的材料中就写了：到 2027 年，训练单个 frontier AI 模型，可能就需要大约 2GW 的联网算力集群；到 2028 年，美国光是 frontier AI 训练总需求就可能达到 20–25GW，而训练加推理合计至少需要 50GW 级别的电力基础设施。
中文
用北约名义，卖武器，要驻军费，买美国货，加关税，要格陵兰岛，真是奸商
AB Kuai.Dong@_FORAB
特朗普在发布会上抱怨，你知道除了北约，还有谁没帮我们吗？澳大利亚、韩国、日本。我们在韩国驻扎了 4.5 万名士兵，保护他们免受金正恩的威胁，而我和金正恩关系很好。 他说过我很多好话。他以前还骂拜登是智障呢。
中文
MOCHA retweetledi
Whilst KeyBanc are very bullish on $AMD, they have turned slightly bearish on $NVDA 😱😱😱
According to KeyBanc, ramp of Nvidia’s latest Rubin GPUs are delayed due to SK Hynix HBM4 qualification issues and Rubin shipments have now declined from 2m to 1.5m!!!! 👀
Because of this delay, ODM production has also now been shifted from June to September. ☝️☝️☝️
The 2025 rumours were right!!! Rubin really was delayed!!!!
They now expect 60k NVL72 racks and 6k Vera Rubin racks this year ☝️
#AI
Jukan@jukan05
Interesting. Rubin shipments in 2026 reduced from 2M to 1.5M units due to HBM4 qualification issues.
English