「天下為公」 - 孫文 一碗暖心美食，芝加哥唐人街 一连4天，看不同的项目，AI 赋能一切，硬件单方面，很多的合作方都是中国制造，中国供应链在智能硬件方面，无可匹敌。 而老外在营销、市场推广、融资能力，story telling 能力在北美市场也是无敌的存在。 美国的机会很多，华人如果能克服文化、语言，主动一点，是有机会。 自己经过了第一天，也换上了销售的性格E模式… 对英式、美式教育长大的港、台年轻人，美国还是较好发展事业的地方。 忙碌了这么多天，去唐人街，吃了越南牛肉粉，这店是广东台山老板娘开的。说实在的，不大行，但服务和说粤语搭救，这乡情陪这半合格“comfort food” 还算满意。 一般在华人餐厅，我都给多点小费，毕竟华人在外乡在唐人街都不容易。也没多少钱，100% tips, 当作祝福给这餐厅支持的家庭。 愿天佑我华夏民族！

有推友说，调整完了…你看反弹了…. 回调一般都是拧毛巾！所以….没这么快，纳斯达克才回调了2.6% “拧毛巾”回调法 双手握住毛巾两端，利用身体重心，一只手向上提、另一只手向下压，像拉大锯一样“不断”扭转毛巾。时而放松，时而拉紧…. 不会一下子就拧完。什么时候拧干？ 要看、观察，毛巾的水分。

“Excessive compression will ultimately lead to a strong rebound.” +Pump! $DRAM $MU

今天的美光 $MU 部分娱乐操作买入和卖出，这样能把持仓成本价，降下来。降成本有助长期屯好资产。 比特币大 bitcoin:native 概率还是会下试7万美金以下。到时候, 会买入吗？！

有推友很“关心”我，问起了我美光的成本价，不用担心，还能抗大幅度的回撤。子弹充足，说实在话，越跌我越开心，让我囤更多优质资产。 $MU $DRAM $TSM $GOOGL

"Markets can remain irrational longer than you can remain solvent." – John Maynard Keynes A widely attributed market adage reminding investors that even if an asset is fundamentally mispriced, the timing of the market's correction is impossible to predict

Merkle3s 每周回顾: BTC 回吐至 $76.8K，现货 ETF 单周净流出近 $10 亿为 1 月以来最大；Hyperliquid Pre-IPO Perp 验证链上定价能力；Base 生态 V4 Hooks + AI Agent 双线回暖。 1️⃣ 宏观：ETF 巨额流出 + 月线收官承压，BTC 回测关键支撑 ➡️ BTC 本周从 $81,700 持续回落至 $76,800，跌幅约 6%。现货

之前连涨6週，是完全亢奋、不理性的 ，未来一个月，拧毛巾进行中，几乎所有股票都会调整，尤其半导体。美光 $MU $SNDK $DRAM 都不例外，不是基本面出问题，而是之前涨的太快，获利盘听风声动起来而已。 美光失守$706支持位，估计有机会下试$572.6-$616。如果跌到这个区间，你会捞底吗？！

Jensen Huang delivered the keynote address at Carnegie Mellon University’s (CMU) on May 10, 2026。 "How hard could it be? Fight for the chance! So run—don't walk. Seize the opportunity for your American Dream in this AI century." - CL Reading the transcripts is better than watching a YouTube video. JENSEN HUANG: "President Jahanian, members of the Board of Trustees, faculty, distinguished guests, proud parents and families, and above all, the Carnegie Mellon class of 2026. Thank you for this extraordinary honor. It is deeply meaningful to be here with Carnegie Mellon, one of the world’s great universities, and one of the rare places that invents the future. Today is a day of pride and joy. A dream come true for you, but not only for you. Your families, teachers, mentors, and friends helped carry you here. Before we talk about the future, thank them. This day belongs to them too. Graduates, please stand up. Stand with me. Come on, you guys. Especially, turn to your mothers and wish them a happy Mother’s Day. For you, this is another step in your life, but for her, this is a dream come true. Please sit down. CMU students like robots take instructions one at a time... To see you graduate from one of the world’s great institutions, this is her moment too. My mom and dad are deeply proud of me as well. My journey is their journey. I am their dream come true. And their dream was the American dream. Like many in this audience, I’m a first-generation immigrant. My father had a dream to raise this family in America. When I was nine years old, he sent my older brother and me to the United States. We ended up at a Baptist boarding school in Oneida, Kentucky. Coal country. A town of a few hundred people. Two years later, my parents left everything behind to join us. They came with little to nothing. My father was a chemical engineer. My mother worked as a maid at a Catholic school. She woke me up at 4 a.m. in the morning to deliver newspapers. My older brother got me a job as a dishwasher at Denny’s. Which, at the time, felt like a major career advancement. That was my view of America. Not easy, but full of opportunities. Not a guarantee, but a chance. My parents came here because they believed America could give their children a chance. How can we not be romantic about America? I went to Oregon State University. I met my wife, Lori, when I was 17 years old. I was the youngest kid in school. We were sophomore lab partners. She was 19. An older woman. I beat out 250 other boys in class and won her heart. We’ve now been married for 40 years. We have two amazing children, both working at NVIDIA. When I was 30, I started NVIDIA with Chris Malachowsky and Curtis Priem, two amazing computer scientists. We wanted to build a new kind of computer, one that could solve problems ordinary computers could not. We had absolutely no idea how to build a company, raise money, or run NVIDIA. I just thought, how hard could it be? It turns out, it is really, really hard. Our first technology didn’t even work. We nearly ran out of money. At one point, I had to fly to Japan and explain to Sega’s CEO that the technology they contracted us to build would not work. Asked to be released from a contract we could not complete. And then asked that they still pay us. Without the money, NVIDIA would vaporize. It was embarrassing, humiliating, and one of the hardest things I have ever done. And Sega’s CEO, Irimaji-san, said yes. I learned early that being CEO is not about power, but the responsibility that comes with keeping the company alive. And that honesty and humility can be met with generosity and kindness, even in business. We used the money to reset the company, and out of desperation, we invented new ways of designing chips and computers that we still use today. For 33 years, NVIDIA has reinvented itself over and over again, each time asking, how hard can it be? And each time learning, it’s harder than we thought. But through those experiences, we learned never to see failure as the opposite of success. Each failure is just another learning moment, a humility moment, a character-strengthening moment. The resilience forged through setbacks is what gives you the strength to go again. Today, I am one of the longest-serving CEOs in technology. NVIDIA, the body of work I share with 45,000 extraordinary colleagues, is my life’s work. Now it’s your time to realize your dreams. And the timing could not be more perfect. My career started at the beginning of the PC revolution. Your career starts at the beginning of the AI revolution. I cannot imagine a more exciting time to begin your life’s work. AI started right here at Carnegie Mellon. Carnegie Mellon is one of the true birthplaces of artificial intelligence and robotics. In the 1950s, researchers here created the Logic Theorist, widely recognized as the first AI computer program. In 1979, Carnegie Mellon founded the Robotics Institute. Artificial intelligence has gone on now to reinvent computing completely. I have lived through every major computing platform shift... But what is about to happen now is bigger than anything before. Because intelligence is foundational to every industry, every industry will change. For the first time, the power of computing and intelligence can truly reach everyone and close the technology divide. Now it’s your time to realize your dreams — and the timing could not be more perfect. AI is making intelligence more broadly accessible... This is the largest technology infrastructure buildout in human history, and a once-in-a-generation opportunity to reindustrialize America and restore the nation’s capacity to build. Every major technological revolution in history created fear alongside opportunity. When society engages technology openly, responsibly and optimistically, we expand human potential far more than we diminish it. AI won’t replace you, but people who use AI will. AI automates tasks but elevates workers. The task and purpose of a job are not the same... AI is unlikely to replace you, but someone who uses AI better might. The responsibility of our generation is not only to advance AI — but to advance it wisely. Scientists and engineers have a profound responsibility to advance AI capabilities and AI safety together. Policymakers have a responsibility to create thoughtful guardrails... So, the answer is not to fear the future. The answer is to guide it wisely. Carnegie Mellon has a motto I love: “My heart is in the work.” So put your heart in the work. Build something worthy of your education, your potential and the people who believed in you long before the world did. We have the opportunity to close the technology divide — and bring the power of computing and intelligence to billions of people for the very first time. To reindustrialize America and restore our capacity to build. And to help create a future more abundant, more capable, and more hopeful than the world you inherited. You are entering the world at an extraordinary moment. A new industry is being born. A new era of science and discovery is beginning... No generation has entered the world with more powerful tools — or greater opportunities — than you. We are all standing at the same starting line. This is your moment to help shape what comes next. So run. Don’t walk."

光模块business timing和market timing - By Frank 他对存储和光模块的对比，把规模解释的很清晰。 现在解答了CPO的scale up 路程时间点。 Frank 是个事实求是的人，看数据、看时间段，不哗众取宠，才华横溢，又朴实谦虚，只有谦虚，才能激起对知识的渴望和求真欲望，而谨慎是严谨求证，validation。 Frank 代表我们每一个普通人，通过深入研究，了解产业链，理清关联逻辑和时间点！这是最重要的，投资里的business timing和market timing 其实很重要。 向Frank学习， 文章值得细琢。

刚开完基金的星期会，预测整个市场下调的开始，纳斯达克下调10% - 13%，半导体平均下调20%左右，到新的一个季度财报季的开始，重新涨回来，纳斯达克再涨20%。 还是要拥抱AI！ 非投资建议，DYOR。

刚开完基金的星期会，预测整个市场下调的开始，纳斯达克下调10% - 13%，半导体15-20%左右，到新的一个季度财报季的开始，重新涨回来，纳斯达克再涨20%。 还是要拥抱AI！ 非投资建议，DYOR。

投资世界里，纪律排第一位 美股（尤其是S&P 500和Nasdaq）确实经历了连续6周上涨，近期出现调整是正常的，需要做好风控。 - S&P 500在5月9日前后连续6周周线收涨，创下2024年以来最长连涨纪录，主要由科技/AI股（如半导体、Nvidia等）驱动， 伴随企业盈利强劲（Q1盈利增长超28%）、地缘风险暂时缓和等因素。 - Nasdaq 和 S&P 500 多次创历史新高（S&P 500 曾接近或超过7500点附近），YTD涨幅约8%，过去一年超26%。 - 近期调整 ：5月中下旬出现回调（如5/15单日跌超1%），属于“高位获利了结+震荡”，很正常- 连续急涨后，市场总有消化获利、关注收益率/通胀/地缘等因素的阶段。 市场本轮上涨强劲（从3月底低点反弹超10-14%），但估值已处高位，短期波动加大是预期之内的。 为什么调整正常？ - 历史规律：连涨多周后，回调概率高（尤其是科技权重大的指数）。没有只涨不跌的市场。 - 当前驱动：AI乐观、企业盈利好支撑长期趋势，但短期有油价、通胀预期、地缘（伊朗相关）、美联储政策等变量。 - 技术面：已近历史高点，获利盘多，容易出现高位震荡或短期回撤，而不是直接崩盘。 风控建议 1. 仓位管理 ：不要满仓追高，留现金或降低杠杆，回调时可分批布局优质资产。 2. 分散风险：别全押科技/AI，可关注盈利稳健的板块，或适度配置防御性资产。 3. 止损/再平衡：设定个人风险承受范围，定期审视持仓。 4. 长期视角：美股基本面仍较强（盈利增长、经济韧性），但短期波动是常态。投资不是赌博，风控永远第一。 总体来说，上涨是事实，调整也正常，做好准备最重要。市场永远有不确定性，保持冷静、纪律执行策略就好。

投资世界里，纪律排第一位 美股（尤其是S&P 500和Nasdaq）确实经历了连续6周上涨，近期出现调整是正常的，需要做好风控。 - S&P 500在5月9日前后连续6周周线收涨，创下2024年以来最长连涨纪录，主要由科技/AI股（如半导体、Nvidia等）驱动，伴随企业盈利强劲（Q1盈利增长超28%）、地缘风险暂时缓和等因素。 - Nasdaq 和 S&P 500 多次创历史新高（S&P 500 曾接近或超过7500点附近），YTD涨幅约8%，过去一年超26%。 - 近期调整 ：5月中下旬出现回调（如5/15单日跌超1%），属于“高位获利了结+震荡”，很正常- 连续急涨后，市场总有消化获利、关注收益率/通胀/地缘等因素的阶段。 市场本轮上涨强劲（从3月底低点反弹超10-14%），但估值已处高位，短期波动加大是预期之内的。 为什么调整正常？ - 历史规律：连涨多周后，回调概率高（尤其是科技权重大的指数）。没有只涨不跌的市场。 - 当前驱动：AI乐观、企业盈利好支撑长期趋势，但短期有油价、通胀预期、地缘（伊朗相关）、美联储政策等变量。 - 技术面：已近历史高点，获利盘多，容易出现高位震荡或短期回撤，而不是直接崩盘。 风控建议 1. 仓位管理 ：不要满仓追高，留现金或降低杠杆，回调时可分批布局优质资产。 2. 分散风险：别全押科技/AI，可关注盈利稳健的板块，或适度配置防御性资产。 3. 止损/再平衡：设定个人风险承受范围，定期审视持仓。 4. 长期视角：美股基本面仍较强（盈利增长、经济韧性），但短期波动是常态。投资不是赌博，风控永远第一。 总体来说，上涨是事实，调整也正常，做好准备最重要。市场永远有不确定性，保持冷静、纪律执行策略就好。

Specifically, Ukrainian drones attacking Russia were electronically hijacked by Russian forces; they subsequently veered off course and flew into the Baltic states. In Latvia, an oil storage facility was struck by an explosion and caught fire. Consequently, the Prime Minister of Latvia has resigned. News from multiple reputable sources (e.g., The Guardian, Al Jazeera, France 24, Reuters, Kyiv Independent, and others).

Specifically, Ukrainian drones attacking Russia were electronically hijacked by Russian forces; they subsequently veered off course and flew into the Baltic states. Consequently, the Prime Minister of Latvia has resigned. In Latvia, an oil storage facility was struck by an explosion and caught fire. News from multiple reputable sources (e.g., The Guardian, Al Jazeera, France 24, Reuters, Kyiv Independent, and others.

未来5年，你认为哪一家Mag 7姐妹公司带来的回报最高?!

两家亚裔创立的POS 公司的融资金额分别是9千万美金和2亿美金，都有个共同点AI赋能！谁说Saas， 没前途？！ 用了产品后，企业客户能轻易离开？！大公司用AI写软件，就能取代？！很难，这样客制化的深入企业资料库的服务。 想想企业级软件股，想想 $PLTR $MSFT 在对的赛道，美国还是黄金满地。

两家亚裔创立的POS 公司的融资金额分别是9千万美金和2亿美金，都有个共同点AI赋能！谁说Saas， 没前途？！ 用了产品后，企业客户能轻易离开？！大公司用AI写软件，就能取代？！很难，这样客制化的深入企业资料库的服务。 想想企业级软件股，想想 $PLTR $MSFT 在对的赛道，美国还是黄金满遍。

今年的全美支付界盛会，在芝加哥的NRA （National Resturant Association) Show。 为什么来这里？看AI怎么落地在Brick and Mortar 场景。这里有新的科技，整合机器人、数据、Omni Channel、crypto/stablecoin payments. AI + Web3 改变我们生活！ 这会有个美国餐饮行业的舆情表作用。比往年来说，人流看似减少了，餐厅行业严重受通膨影响，人工成本（业界人士透露ICE抓人也是主因之一），加上油价，大家都减少出外就餐。 在马路上也很少看到电动车… 这“油价”影响生活，肯定影响选情。 尽管如此，还是预期5/16-19号，有超过5-6万的参会人员…

在AI产业中，这是关乎“水与空气”的问题 2027年，LPDDR组件的消耗量将迎来爆发式增长！ - LPDDR（亦称LPDRAM）即“低功耗双倍数据速率存储器”。 接下来应当提出的问题是：“那么2028年和2029年又将如何？” 您的答案会是什么？ $DRAM $MU $SNDK It's about "Water and Air" in the AI industry LPDDR components​ consumption Pump + in 2027! - LPDDR, also known as LPDRAM stands for “low-power double-data-rate memory” The next question that should be asked is, "How about 2028 and 2029?" What would your answers be?

How about 2028 and 2029?! "Expect the following markets to represent >80% of 2026 WFE market growth, similar profile in 2027 - Leading-edge F/L - DRAM - Advanced packaging"

拿了行李，坐上Uber，这不是大麻的味道吗？！心想这司机没抽吧？！也许是上一个乘客留下的“自由味道”？！ 快一年没到美国来看看了，对这美丽国的味道真不惯！让司机把窗户打开。 太累了，在车上打个瞌睡…. 放下行李，就要去开会…

拿了行李，坐上Uber，这不是大麻的味道吗？！心想这司机没抽吧？！也许是上一个乘客留下的“自由味道”？！ 快一年没到美国来看看了，对这美丽国的味道真不惯！让司机把窗户打开。 太累了，在车上打个瞌睡…. 放下行李，就要去开会…

难道日本第一投行，Nomura 野村证劵也看错了！要给三星和海力士Re-rate， 不再用如此不合理低PE倍数，因为AI super cycle，供不应求，寡头垄断，技术护城河？！还是这帮分析员看到确实的改变？！

AI 产业护城河最宽、最深的企业是哪一家？

别在意，一切皆为交易！The Art of the Deal II 川普以「好莱坞男星」比喻习近平，形容气场强、没人更合适做他这位子。身材外貌高大，有领袖气息；而行事风格，专注大局，不要花招！ 川普： “习近平为伟大且备受尊重的领导人、朋友。” 川普的却是个出色的商人总统，很懂内地那一套！ 能屈能伸，这样大豆能卖了、加上石油、波音飞机、GE 发动机，业绩不错；再卖点烂铜烂铁给绿蛙政权，这样也不得罪习，然后，把霍尔木兹海峡处理一下，股票市场先洗一洗，7/8月股票再开始冲，尝试拿下中期选举，至少别输的太难看。 川普讲究的是里子、实惠。 为国、为民谋利，这一点，他的却做的不错。 值得敬重。

别在意，一切皆为交易！The Art of the Deal II 川普以「好莱坞男星」比喻习近平，形容气场强、没人更合适做他这位子。身材外貌高大，有领袖气息；而行事风格，专注大局，不要花招！ 川普： “习近平为伟大且备受尊重的领导人、朋友。” 川普的却是个出色的商人总统，很懂内地那一套！ 能屈能伸，这样大豆能卖了、加上石油、波音飞机、GE 发动机，业绩不错；再卖点烂铜烂铁给绿蛙政权，这样也不得罪习，然后，把霍尔木兹海峡处理一下，股票市场先洗一洗，7/8月股票再开始冲，尝试拿下中期选举，至少别输的太难看。 川普讲究的是里子、实惠。

投AI，投主旋律，但最后的一切都交给市场，敬畏市场！ 这一次，聊聊存储和光模块。 无论光模块或者存储，只用以前的眼光来看都不合适，光模块是插件模块加新的机会CPO， 存储是（DDM +NAND）+ HBM，其实还有很多其他例如HDD等， 但都需要不断迭代服务使用者。 什么标的，到最后其实都是要看估值，看兑现业绩。 HBM现在就已经是看到的大出货量，seeing is believing 眼见为实，未来的增长需求只会不断增加，而供应没法赶上，建厂到良率提升都需要时间，24个月是比较靠谱的。工艺，也就那三家。大陆的生产商，还有很长时间才有可能进入英伟达的供应清单里。CPO 真正交付到什么时候开始？看看LITE 就知道。 而光的毛利率是多少 ？40多%对比70多% （存储），美光的新季度毛利会攀升到80%以上了。有人认为HBM是商品commodities， 那光模块就不是？其实某程度上都是。 其实都没有错对，最后看市场。而且无论存储和光模块的股票，最重要的两点，除了选股外，就是买入的时机，早买入成本低，但必须有一定的确定性（还有资金的体量重仓而否），余下就是持有的时间和退出，就是如此简单。 技术就是护城河，毛利就是最好的KPI，也是供需的决定是否是卖方市场。 其实都是靠英伟达 Nvidia吃饭，LITE  2027 会出貨 CPO 給 hyperscaler 的量小、单价非常高、毛利（对比HBM还是逊色不少）很好看。但大規模取代估计2029- 2030，不会是明年。 而且暂时看总体量远低于存储。有一天，云服务商、大模型供应商不花钱了、减少投入了，大家（存储、光模块）都要慌了。 值不值得买，还是那两个字“估值”，公允值，back to square 1, numbers DCF。 存储用PE没问题，但要分产品线，大行如摩根等也用了这方法，HBM， DDM, NAND分开，只要业绩持续，自然会调高目标价。 其实，两者都值得学习，尽量看透。 至于投哪些股票，的却很多时候都有一定的主观性，但有一点，还是要投在主旋律上！仓位还是需要分散一点。 一切交给市场，敬畏市场！ $DRAM $MU $SNDK $LITE $COHR $AOOI $GLW

钱，只会流向开放经济体。 很久没远行了，要补课去了！ HKG-> Chicago -> NYC -> SFO ...... -> HKG 对于普通人，越开放的地方，机会越多。 很久没出远门了，离开自己生活的地方几个星期，有机会，以外人视角审视自己身处的大环境。避免困在信息茧房。用眼和耳朵，看大趋势的改变。 在一个封闭的地方，普通人难有大机会。而开放的地方，普通人有机会通过技术、智慧、勤奋，改变人生。 去年十月，从新加坡搬回香港，然后，在内地出差比较多… 深深的感受到，国内经济的不振，而香港经济在改善中。 越开放的地方，机会越多。 钱，只会流向开放经济体, 越锁国，越贫困。

哪个航空公司的空姐最漂亮，是你的梦中情人类型？如不在选择里，欢迎评论区留言分享，让大家见识一下大千世界。

咱们幽默一番，如果你坐飞机十几个小时，你会干什么？

现在的却是时候存储部分现金“积谷防饥”，候机抄底，AI主线骨干股票，你认为的“美股AI主线骨干”标的是哪些？

The TSMC chart shows their System-on-Wafer (SoW) tech scaling up interposers massively. By 2026: ~12 HBM stacks per substrate. By 2029: 64 HBM stacks (5x jump). It means next-gen Al accelerators/GPUs will pack way more high-bandwidth memory. Demand for HBM/DRAM (and $MU , $TSM ) keeps exploding—not peaking. Al hardware is just ramping.

量子 2026：拐点已至，信仰过热! 量子计算 2026 年是真实的产业拐点，但含 NASDAQ 量子标签这件事的稀缺性溢价，已经被市场提前定价了。 五年之后回望，你会发现我们今天讨论的不是量子计算公司值多少钱——而是美股量子标签值多少钱。 这两个问题，答案完全不同。

谦逊、敬畏市场，预备应对“持续”大波动！ 昨晚的大跌是个市场给大家的提示，是时候审视组合，做适当回应。 只有零杠杆，才能在大波动市场，活下来！切记，尽量不要用杠杆，真的用也是低杠杆如10%。 重温查理·芒格的箴言 “破产有三种方式 ： 酒精、女人和杠杆。”

自己三个组合的现金仓位分别是40%，20%，20%。 要出差一段长时间，这样能做到心安、眠安、家安，进退自如。坦率的说，还是执行的不够好。 另外一个合伙人管理的组合现金是50%，YTD回报45% , 他做的更好，更丝滑，目标回撤也会控制在10%以内。

谦逊、敬畏市场，预备应对“持续”大波动！

自己三个组合的现金仓位分别是40%，20%，20%。 这样能做到心安、眠安、家安。 My cash positions in the three portfolios are 40%, 20% and 20% respectively. This allows me to have peace of mind, sleep soundly, and enjoy peace at home.

自己三个组合的现金仓位分别是40%，20%，20%。 这样能做到心安、眠安、家安。