$HIMS long term shareholders just realized Short sellers are abt to lose billions at $50-$70-$100-$150-$200. Current Short Interest: 54,945,417 shares Short Percent of Float: 26.69% 🐻better be ready! Not Financial Advice! DYOR!

$AMD | The real reason why @AMD outperforming all Semis because it was misunderstood🧐🧵 AMD is still very underowned among Retail investors and Active Funds. Because they didnt take CPUs and GPUs share gain seriously. AMD Is projected to get 15-25% GPU market share (FY2027) and Dr. Su aims for 50%+ for CPUs market share. Not that long ago, months ago, AMD was the least owned Semi stock out of all Semi companies. The real Institutional FOMO will be in November when we get Q3 ER where Q4 guidance is issued. AMD has some of the smallest public float or effective float. I dont want to hype u up, but there isn't much supply. Stock price appreciated so much since March because of this very reason. I covered the most recent AMD Conference for subscribers in August and September 2026, where u get the lastest data. I said repeatedly in 2024-2025, we will see analysts upgrading weekly/monthly on AMD, and I was right. AMD was mispriced for its potential, especially in the world of Inference and Agentic AI. Inference is now more than 60% of AI compute buildout vs training, and Inference is growing extremely FAST! I will write a thread on AMD PEG Ratio later, and I believe this is more accurate to judge AMD vs Fwd PE or how we try to project different FY2027 EPS. Alright that is it. Not FInancial Advice! DYOR!

$AMD shareholders just realized, if @AMD were to trade at $INTC's valuation or Fwd P/E of 65-100x, AMD would trade at $1,400-$2,200 per share.

Why Wall Street thinks $AMD is biggest Consumers Agentic AI winner 🧵| $META @Muse $MSFT $AMZN $GOOGL @AnthropicAI @OpenAI Not Financial Advice! DYOR! Consumer agentic AI is not another GPU story. A personal agent like Muse is a worker with a computer: a private VM, a real browser, files, logins, background jobs, and the ability to keep working after the user closes the app. The model still needs accelerators. The new demand spike is the army of small, always on computers those agents live in. That work runs on CPUs, and the company best built for it right now is @AMD . The product that wins consumers is the agent that can use the world people already use websites, checkout, email, documents, and a long tail of software that was never rewritten for a new chip. That world still speaks x86. AMD’s high core EPYC parts add the other thing this workload cares about: thread density. One Venice class socket can run 256 cores and 512 threads, so a rack can host more agent sandboxes without waiting for software to be ported. $ARM can be slightly cheaper when a hyperscaler owns the guest image and only needs Linux plus its own browser. It is not a better consumer computer. That is why the cleanest Wall Street call is not “all CPUs win equally.” UBS’s framework is the honest one: standalone agentic racks look more like 60% x86 and 40% ARM, and AMD is the x86 vendor in position to take the extra demand. Bank of America just raised AMD to a $720 target on the same idea, agentic work turns the CPU from a supporting actor into a growth engine, and AMD covers both the dense sandbox chip and the fast host next to a GPU. Intel benefits because it is the other x86 name, but it is not the leader. That is why hyperscalers and AI labs cannot just scale consumer agentic on a pure ARM bill of materials. They will keep buying Graviton, Cobalt, Axion, Vera, and Arm AGI for orchestration, efficiency, and GPU-adjacent host nodes they fully control. The moment the agent has to live in a real sandbox compile code, drive a browser against live sites, run mixed tools, spawn thousands of concurrent VMs, they need the x86 density and software stack EPYC already ships. Meta, OpenAI, Anthropic, and the clouds can mix architectures. They still have to add significant EPYC capacity (it is projected to be 50-60m units per year combined) if they want consumer agents that act like people on the open web, not only on a company built island. That mix is why AMD is the biggest CPU winner because it has the product, the share gain flywheel, and a second engine. Server CPU revenue is already exploding, EPYC Venice is purpose built for agent concurrency, and Ryzen AI keeps AMD in the on device Windows PC where a consumer agent wants to touch local apps. On top of that, Instinct and Helios mean AMD is not betting the company on CPUs alone. When Meta, enterprises, and cloud buyers buy agent infrastructure, AMD can show up in the sandbox, the host node, and the accelerator tray. Consumer agents multiply CPUs. The CPUs that can run the open consumer stack at massive concurrency are high thread x86. AMD is shipping that chip now, taking share, and pairing it with GPUs instead of praying the CPU cycle is the only cycle. If personal agents scale from a novelty to a default layer of the internet, AMD is the name that sits at the center of that buildout. Dr. Lisa Su has been saying this out loud. On the earnings call she told analysts the old 1 to 4 or 1 to 8 CPU to GPU host ratio is moving toward 1-2 to 1, and “you can even imagine if you get lots and lots of agents that you could have more CPUs than GPUs.” At Advancing AI she went further: Venice “is actually designed for the agentic era,” and “EPYC is the only CPU portfolio that leads across all three” roles; GPU host, dense agent sandbox, and general purpose enterprise. Consumer agents multiply CPUs. The CPUs that can run the open consumer stack at massive concurrency are high thread EPYC. AMD is shipping that chip now. Not Financial Advice! DYOR!

$AMD shareholders just realized there will be intense comeptition with $META @Muse from $AMZN $GOOGL $MSFT $AAPL @AnthropicAI @OpenAI to perform Agentic Tasks for consumers(B2C). And @AMD has the best x86 CPUs to truly scale Consumers Agentic. The biggest Agentic AI use case at massive scale to generate Hyper Revenue growth for decades to come. This is a massive AMD CPU shortage that most will likely see a bidding war. Meaning AMD will have ~Pricing power ~Margin Expansion ~Years of TAM & Market Share Capture and Expansion ~Even faster Revenue Growth ~ Much higher multi-year CAGR than what was projected just weeks ago This is going to be fun! AMD shareholders gonna do VERY WELL in the next 3-5 years! Not Financial Advice! DYOR!

BREAKING $AMD hit new ATH today and broke my $620 Old Personal Price Target 🚀🚀🚀 Disclaimer: I do not know or see the future. Just pure fundamental researches. Im still keeping my personal PT on @AMD to $800 by end of 2026. 15-25% GPU market share in 2027 and biggest Agentic AI winner with superior CPUs(EPYC Venice) If there are anything material change such as ~Massive deals with $GOOGL and $AMZN ~More CPU exclusive multi-year deals ~Massive increase from TSMC Allocation I reserve the right to change my personal Price Target! Congratz to all AMD OG shareholders. For bears and short sellers that bet against me. May be move on to other tickers that dont have my name on it. Not FInancial Advice! DYOR!

$AMD $PLTR $TSLA $XYZ $GRAB $HIMS $META $GOOGL $AMZN ... I'm very selective when it comes to long term winners. Not all companies met all my ✅ But, before that, I put in 500-1000+ hours to study these businesses. Most companies did not make it from the first 20-50 hours. Then I find the "secret sauce", leadership and long term thesis. I prefer strong and durable fundamental over hype, most of my long term were bought at extreme compressed valuation or ATL. The combination of obsessive hours, secret sauce , founder/CEO quality, fortress balance sheet+ operating leverage, and refusal to sell winners is how I win versus typical growth or momentum screens. And I dont sell winners! Please do your own DD before buying/selling. My risk tolerance and duration are different than yours Sincerely, MikeLongTerm Not Financial Advice! DYOR!

If $AMD reaches my $800 Personal PT before Q3 ER or November, that would mean institutions expect $20B Q4 guidance from Q3 ER. So roughly 100% YoY. As soon as Q4 guidance is out, even at $800, we would be looking at 26-28x Fwd PE. Pretty reasonable fwd P/E for this kind of growth IMO. It is going to be fun to watch folks arguing what I laid out in 2024-2025. The CPU Ramp will start from Q4 2026, but Dr. Su will have to secure additional 30% more Wafers for CPUs to service $META @Muse . CPU supply is so far behind that even if TSMC doubled CapEx - AI Labs, Enterprises and Hyperscalers would still be supply constrained. Not Financial Advice! DYOR!