$Palantir Technologies Inc.(PLTR)$ As $PLTR shareholders know, pessimists sound smart, optimists make money.

$美国超微公司(AMD)$  $AMD is currently disrupting $NVDA and will take a large % of its marketshare in AI for 2 key reasons: 1. Declining CUDA Lock-in: NVIDIA's CUDA software has been a major factor in their success, but its grip is loosening as PyTorch, a more versatile framework, gains traction. AMD GPUs are now compatible with PyTorch, making them a more attractive option for developers. 2. Chiplet Architecture Advantage: AMD's chiplet architecture enables them to produce high-performing GPUs at lower costs. While chiplet-based GPUs initially have lower margins, AMD's iterative approach suggests they will eventually outperform NVIDIA in both performance and price. NVIDIA's reluctance to embrace chiplets could prove costly as AMD gains momentum. In

$美国超微公司(AMD)$  Here's why I think $AMD is quite cheap at $158. $NVDA's price to sales ratio is 31.24 and $AMD's is 7.51. For income statements that are not in a phase of maturity, the price to sales ratio is a measure of enthusiam. The above spread is bound to close going forward and I believe the P/S ratios of the companies will eventually come to parity. In absolute terms, I believe both companies have a long way to go as the world demands exponentially more computation over the coming decade/s. But in terms of valuation, the market is currently valuing Nvidia as the sole provider of AI chips when in fact: 1. $AMD is a strong contender to a number 2 position, as a provider of AI chips. 2. It is very likely that $AMD's hardware can actually

$美国超微公司(AMD)$  Everything you need to know about $AMD, in 8 bullet points. 1. $AMD's rise over the last decade has been the result of the company betting on chiplets. Chiplets consist in making a chip out of many small chips, instead of making a big one (monolithic) directly like $NVDA does. This is how it has disrupted $INTC and how it is now poised to disrupt $NVDA. 2. We are nearing the "end" of Moore's Law. Making monolithic chips is getting exponentially harder and therefore, chiplets are the way around the end of Moore's Law. They enable much higher yields at similar if not higher levels of performance. $AMD has been working on chiplets for about a decade now, while the rest of the industry is only just catching up. 3. $AMD is just a bun

$英伟达(NVDA)$  It won’t be so easy for $AMD to disrupt $NVDA, because $NVDA’s GPUs have massive network effects. Just as $TSLA vehicles share a unified architecture, $NVDA's GPUs are interconnected by CUDA, a cohesive software framework that fosters a self-sustaining ecosystem fueling the company's growth. CUDA acts as the bridge that seamlessly integrates $NVDA's diverse GPU lineup, enabling developers to harness the power of these computational workhorses. With each software iteration, the network of compatible GPUs expands, attracting a growing pool of expertise and talent. The broader the adoption of $NVDA's GPUs, the more valuable they become, solidifying the company's position as a dominant force in the GPU market. $NVDA's commitment to s

$Palantir Technologies Inc.(PLTR)$  AI is forecast to quickly transform the healthcare industry, aiding in analytics, diagnosis, data management, communication, and drug discovery. The AI in healthcare market is expected to reach $208 billion by 2030, expanding at a 36.4% CAGR from $32B in 2024.

$Hims & Hers Health Inc.(HIMS)$ Much of $HIMS 100X potential comes from its unique approach to AI in the healthcare space. In Q3 2023, $HIMS launched Medmatch, a “proprietary service that deploys artificial intelligence and machine learning against the expansive dataset at the core of the Hims & Hers platform.” When comparing $HIMS architecture to $TDOC, $HIMS can deploy this service for each of its verticals and translate the insights it yields into productive modifications of the healthcare process–without having to deal with the inflationary, slow, and costly traditional healthcare system. With nearly 1.5M subscribers–and assuming $HIMS has been storing the anonymized data–MedMatch will see in one day more patients than a physician can

$UiPath(PATH)$ is now set to print cash. In my original UiPath deep dive, I pointed out the two key drivers that will potentially trigger a financial inflection point: 1. Semantic automation: this will allow UiPath to better understand the workflows and their content, in order to automate tasks of higher value over time. 2. Low code: this will allow UiPath to democratize access to the platform, by enabling anyone to automate workflows. (Low code enables folks to build things without code). This quarter I see no particular progress made on the semantic automation front, while UiPath is advancing well on the low code front. In Q1 2024, UiPath released Clipboard AI, which I identified as the key mechanism via which UiPath was developing semantic auto

$CrowdStrike Holdings, Inc.(CRWD)$ $CRWD is looking like a winner takes all candidate. The company has one of the strongest moats in the XDR space in cybersecurity, which is enabled by the following 5 dynamics: 1. Lightweight Agent: Easiest deployment & management, attracts customers, generates more data. 2.Unified Data Model: Unrivaled threat landscape view, superior AI model training. 3. Cost-Effective AI Modules: Individually trained on customer data, high operating leverage. The cost of deploying additional modules is marginal. 4. Self-Reinforcing Growth: New modules attract customers, generate data, improve AI and lead to the creation of even better modules. 5. Fixed Cost, Scalable Revenue: More modules per customer, significant profit

$英伟达(NVDA)$  TSMC’s CoWoS capacity is expected to rise from 15,000 at the end of 2023 to 17,000 by the end of Q1, before expanding through the rest of the year. It is estimated that half of TSMC’s CoWoS capacity is allocated to meet Nvidia’s demand.

$美国超微公司(AMD)$  $AMD is already disrupting $NVDA on the hardware side. $AMD is employing the same strategy that once propelled it past $INTC. Now, $AMD is poised to disrupt $NVDA's stronghold in the GPU market, utilizing its chiplet-based architecture to challenge the monolithic approach that has been $NVDA's hallmark. $AMD's chiplet-based design breaks down a large chip into smaller, more manageable modules, known as chiplets. This modular approach offers several advantages over $NVDA's monolithic architecture: 1. Higher Yields: If a single chiplet fails during manufacturing, the entire chip does not need to be discarded. This results in significantly higher yields and lower production costs. 2. Scalability: $AMD can easily add or remove chip

$Palantir Technologies Inc.(PLTR)$  The market now understands $PLTR's importance in the AI industry. But it is yet to figure out that $PLTR is emerging as a dominant cloud player, positioned to disrupt the uninterrupted distribution of AWS, $GOOG Cloud and more. While established players like $AMZN, $GOOG, and $MSFT offer general cloud infrastructure, $PLTR is taking a different approach to become a key player. They're moving beyond the "raw compute" model, where users build their own solutions, and instead offering pre-built, tailored solutions that address specific business needs. Think of it as: A. Traditional cloud: Buying building blocks (infrastructure) and building your own solutions (time-consuming, expertise required). B. $PLTR's a

$Palantir Technologies Inc.(PLTR)$   is a trillion dollar company in the making.

$美国超微公司(AMD)$  Everything you need to know about $AMD before today's Q4 earnings, in 8 bullet points: 1. $AMD's ascent over the past decade can be attributed to its bold bet on chiplets, a revolutionary approach to semiconductor design that stands in stark contrast to the monolithic approach favored by industry giants like $NVDA. By breaking down complex chips into smaller, interconnected modules, $AMD has managed to achieve higher yields, improved performance, and reduced costs, paving the way for its resurgence in the semiconductor market. 2. As Moore's Law nears its limits, manufacturing increasingly complex monolithic chips has become increasingly challenging. Chiplets offer a viable solution to this challenge, enabling $AMD to continue pu

$Snowflake(SNOW)$ Earnings Snapshot:

$特斯拉(TSLA)$  TSLA is on the verge of a major breakthrough, one that will dwarf all previous ones. The company's manufacturing prowess continues to improve fast, enabling it to accelerate its AI and energy businesses. The combination of these three businesses will create a platform akin to the internet.  $TSLA's core competency lies in optimizing its unit economics., which is now spilling over into other areas: renewable energy and AI.  The better $TSLA gets at manufacturing, the better its production of new battery packs and solar panels becomes. The more cars $TSLA manufactures, the more data it can gather for training its Full Self-Driving (FSD) software. Two metrics in particular have shown an inflection point since Q2 2023: 1. C

$Palantir Technologies Inc.(PLTR)$  I think $PLTR will eventually own the cloud compute market and become a trillion $ company, driven by its groundbreaking approach to valuable compute. In the current industry, most companies acquire raw compute, which they then customize by programming software to fit their specific needs. However, $PLTR is breaking new ground by transitioning to valuable compute, positioning itself as a crucial player in the cloud realm. Through its digital twin generation for Company N, $PLTR creates a blueprint of repeatable infrastructure that becomes accessible to Companies N+1 and beyond. This transformative approach liberates these subsequent companies from purchasing raw compute and enables them to acquire comput

$Hims & Hers Health Inc.(HIMS)$ $HIMS is a company with 100x potential. $HIMS is a tele-health company that connects patients with physicians via an app and fulfills medication via automated pharmacy facilities. These are the 6 core value drivers of the company: 1. A cost advantage: $HIMS can treat a growing range of conditions at a cost below what patients would pay out-of-pocket via their insurance. This is a strong moat in an industry that is otherwise an inflationary hamster wheel. 2. Expanding verticals: Every quarter $HIMS is treats more and more conditions. Management has demonstrated great discipline in picking new verticals, with the company increasing its cash from operations. 3. Extraordinary organizational properties: $HIMS has

$美国超微公司(AMD)$  The market is processing $AMD earnings wrong: 1. The market is looking at the AI story as if it were only about selling GPUs. But $AMD's business segments are effectively distribution channels via which it can repackage and sell its core AI tech. $AMD is going to be adding AI functionality to all of its products. Over the long run, this is a much better strategy than only going head to head with $NVDA selling GPUs (which it is doing too), simply because it makes distribution easier and cheaper. The datacenter growth is being offset by cyclical market weakness in gaming and embedded markets, but these two segments will give $AMD an advantage over time. "[...] we see clear opportunities to drive our next wave of growth as we de

$美国超微公司(AMD)$  $AMD's Lisa Su key quotes during the Q4 call: "Customer response to MI300 has been overwhelmingly positive." "The additional functionality and optimizations of ROCm 6 and the growing volume of contributions from the Open Source AI Software community are enabling multiple large hyperscale and enterprise customers to rapidly bring up their most advanced large language models on AMD Instinct accelerators."