Share your thoughtsWhat's on your mind?
Fantastic … Thank you.
A user distilled the investing ideas in 417 community posts into 62 knowledge cards, uploaded them into a LongbridgeAI knowledge base and wired it to an Agent — which now checks those notes first rather than leaning on the model's general knowledge.
A user distilled the investing ideas in 417 community posts into 62 knowledge cards, uploaded them into a LongbridgeAI knowledge base and wired it to an Agent — which now checks those notes first rather than leaning on the model's general knowledge.
0818 | Dolphin Research Focus: 🐬 Macro/Industry
1) The 30Y UST yield climbed to 5.32%, the highest since 2007, with the 10Y nearing 5%. The breakdown in U.S.-Iran talks fueled a 3-day oil rally, while global stocks and bonds weakened in tandem and Japan fell over 2%.Large fiscal deficits, oil-driven inflation, and heavy AI infra issuance are pushing up the long end. Equity multiples and risk appetite are under pressure.Higher long-end yields compress valuations for long-duration tech growth. Geopolitics are supporting oil and reinforcing sticky inflation, leaving risk assets under near-term pressure. Financing demand for compute and AI infrastructure continues to ramp...A TON OF THINGS HAPPENED IN THE STOCK MARKET TODAY.
Here's a full recap:1. Nvidia $NVIDIA(NVDA.US) is backing OpenAI’s Ohio AI campus with a $1.5B investment in SB Energy and support for an initial 4.25GW of AI infrastructure at the PORTS-Pike campus. The project includes an option to expand to the full 8GW. SB Energy will build and operate the site under a 20-year OpenAI lease, with capacity coming online in phases starting in 2028. SB Energy and SoftBank also plan at least 10GW of new power generation and $4.2B of regional grid investment tied to the buildout.2. Anthropic’s revenue run rate has reportedly surged ahead of its IPO, rising to more than $65B in July 2026 from $47B in May and $9B at the end of 2025. The company also posted over $11.5B in preliminary revenue for its latest completed quarter, up from $787M a year ago, while generating positive adjusted operating income. Anthropic has confidentially filed to go public and could debut on Wall Street as soon as this fall.3. U.S. interest expense on the national debt has reached a record $1.4 trillion over the last 12 months, with debt-servicing costs nearly tripling since 2020. If rates stay near current levels, interest payments are projected to rise to $1.7 trillion by November 2028, putting them on pace to overtake Social Security as the government’s largest expense for the first time. The move comes as the 30-year yield hits its highest level since 2007, while the 10-year Treasury yield has crossed 4.7%.4. Evercore ISI’s Julian Emanuel says the $SPY S&P 500 could reach 9,000 over the next 12 months, while maintaining a 7,750 base-case target. He argues the usual bull-market killers — recession, sharply higher long-term yields, and extreme investor chasing — still have not shown up. At the same time, 121 S&P 500 stocks now have negative beta to the index, the highest number since the 2000–2001 dot-com unwind, highlighting how unusual market internals have become.5. Fabrinet $Fabrinet(FN.US) reported Q4’26 revenue of $1.32B, beating estimates of $1.27B and up 45% YoY. Adjusted EPS came in at $4.10 versus $3.82 expected, up 55% YoY, while operating income was $134.25M and net income reached $139.3M. For Q1, Fabrinet guided revenue to $1.38B–$1.43B, ahead of estimates of $1.32B, with adjusted EPS of $4.10–$4.25 versus $3.96 expected. Management said the company delivered record quarterly revenue above its guidance range and remains optimistic about the strength of the business and durability of its growth trends.6. The top 10 most active options today by contracts traded were $NVIDIA(NVDA.US) with 2.8M contracts, $Tesla(TSLA.US) with 2.1M contracts, $Micron Tech(MU.US) with 1.3M contracts, $SpaceX(SPCX.US) with 1.1M contracts, $Apple(AAPL.US) with 1.0M contracts, $Amazon(AMZN.US) with 980K contracts, $Meta Platforms(META.US) with 964K contracts, $Microsoft(MSFT.US) with 849K contracts, $Intel(INTC.US) with 751K contracts, and $AMD(AMD.US) with 572K contracts.7. AI video startup Higgsfield raised a $400M Series B at a $5.4B valuation, with investors including Goldman Sachs, Intel, and DST Global. The 2-year-old company’s annualized revenue reached $700M in August, up from just $20M a year ago, while its user base has grown to more than 30M across 238 countries and territories. Most revenue now comes from businesses using Higgsfield’s AI tools to create marketing content.8. Tesla $Tesla(TSLA.US) is reportedly preparing to launch its purpose-built Cybercab in Austin as soon as this month, according to The Information. The two-seat robotaxi has no steering wheel or pedals, with Tesla planning to start with employee rides on public roads before adding Cybercabs to its Austin Robotaxi service days later. Tesla began Cybercab production in Texas last month and is continuing testing while training local first responders ahead of the rollout.9. Uber $Uber Tech(UBER.US) is investing in Zipline as the companies expand drone delivery for Uber Eats. The goal is to reach 1M drone deliveries per day by the end of 2029, with Uber expecting drones to enable faster deliveries over longer distances. The company sees drone delivery becoming a meaningful growth driver for Eats as it pushes deeper into autonomous logistics.10. China’s credit data weakened sharply in July, with net new loans falling by $50.4B, only the third monthly decline this century and more than 3x worse than expected. Lending to the real economy was even weaker, with net repayments of $87.5B, the largest monthly drop in records going back to 2002. While aggregate financing still rose by $207.7B, nearly all of the increase came from $192.9B in government bond issuance rather than private-sector borrowing, pointing to soft corporate investment, weak household demand, and continued pressure in property.11. Morgan Stanley sees Amazon $Amazon(AMZN.US) with a bull-case path to $500/share by year-end 2027, driven by AWS potentially scaling toward $1T in annual revenue over the next 8–10 years. Analyst Brian Nowak says that scenario could support roughly $500B of company-wide EBIT, while the firm’s base-case price target remains $335. Morgan Stanley argues Amazon’s $1T AWS vision reinforces the size and ROIC of the AI infrastructure opportunity, with AWS currently around $170B annualized and management seeing AI margins and returns tracking similarly, or even slightly ahead, of where core AWS was at the same stage. The key constraint remains compute capacity, with Morgan Stanley estimating Amazon can bring on 6–8GW of compute capacity in 2026/2027 and potentially add around 8GW per year going forward if execution continues at pace.12. UBS expects Nvidia $NVIDIA(NVDA.US) to beat FQ2 revenue estimates by roughly $3B–$4B, with revenue reaching around $94B–$95B. For FQ3, UBS sees guidance in the $107B–$108B range and believes revenue could ultimately exceed $110B. Analyst Timothy Arcuri says Blackwell demand remains stable, while Rubin units are starting to layer in ahead of a bigger FQ4 step-up as Rubin sell-in accelerates toward roughly 500,000 GPU units per month and Blackwell begins winding down. UBS argues the numbers matter more than the AI infrastructure narrative, and expects investors to gain more confidence in a path toward $15+ EPS in C2027 and $20 EPS in C2028. The firm also says memory-driven capex inflation means compute supply is still falling short of demand, which could support another major backlog step-up on the earnings call.WALL STREET IS THE GREATEST SHOW ON EARTH.Source: amit
Taiwan power semiconductor makers plan to hike prices 10%-15% on chips not-under-contract by October amid hot AI demand and after industry giant Infineon raised prices, media report. This is the 3rd time Taiwan firms have raised prices, up an average 15%-20% in the 1st half of 2026. Demand in the 2nd half has further strengthened.
Nvidia’s AI Factory plans call for an upgrade to 800VDC, raising performance demands for all components and opening new growth for high end power chips made using SiC and GaN materials (Silicon Carbide and Gallium Nitride). $Infineon Technologies(IFNNY.US) $STMicroelectronics NV(STM.US) $Texas Instruments(TXN.US) $ON Semiconductor(ON.US) $Wolfspeed(WOLF.US) #FujiElectric #Rohm #MitsubishiElectricSource: Dan Nystedt
When Nvidia announced Spectrum-X switches built with CPO (co-packaged optics) have entered mass production, it named 4 Taiwan firms among partners, TSMC, Foxconn, ASE and Browave as well as China’s leading optics firm, TFC Communication, media report, and multiple firms say their order books are already full through 2027 and expect supplies to be tight into 2028.
CoreWeave, Lambda and Oracle will become the first to deploy the 200Gbps/lane CPO Ethernet switch system, billed for a big leap in performance:5x improvement in network power efficiency5x increase in continuous AI uptime10x extension in Mean Time Between Failures (MTBF)Up to 1.6x overall AI network performance boost, supports 100s of 1000s of GPUsTSMC and Foxconn are expected to play key roles, with TSMC leading production via its COUPE silicon photonics engine platform, which integrates components that are then packaged by ASE and use Browave’s optical components. Foxconn does final assembly of the CPO switches.Foxconn Chairman Young Liu has said CPO switch shipment volumes will reach 10,000 units this year, with 2027 shipments multiplying several times over.Taiwan’s FOCI and ShunSin have secured key niches in the supply chain, with FOCI’s expertise in Fiber Array Units (FAU) and ShunSin’s advanced packaging mass production prowess, media say, calling them key winners further down the supply chain. $NVIDIA(NVDA.US) $Taiwan Semiconductor(TSM.US) $Advanced Semiconductor Engineering(ASX.US) $Oracle(ORCL.US) $Coreweave(CRWV.US)Source: Dan Nystedt
GPU fleet unit economics.
TLDR: Tough to be bearish on $Nebius(NBIS.US) or $Coreweave(CRWV.US) right now.Some maths, but it's important to understand.Using $Nebius(NBIS.US) 1-3yr contract economics:- $20-25M ACV per MW at a payback of 1 year 10 months.- Take $22.5M midpoint ACV/MW at a ~55% cash margin ( $Nebius(NBIS.US) core AI cloud runs ~50%, $Coreweave(CRWV.US) 59%) and you get ~$12.4M/MW of annual cash flow. (Payback runs against cash flow, not revenue)- 1.83 years of that implies "all-in" capex of ~$20-23M per MW (GPUs included). - Which ties to Nebius' $20-25B capex guide for 2026 against ~1GW.- Over a 4-5 year life that's roughly $50-62M/MW of cumulative cash flow against ~$22M of capex. - NPV is +$20-25M per MW even discounting at 12-15% and IRR near 50%.Then years 1-2 of DCF nearly cover the entire capex on their own. Meaning that the GPU could go to $0 residual after year 2 and the deal basically breaks even. That is insane.So really, the bear case doesn't need slow decay to be wrong. Rather, it needs the contracted years themselves to fail.And those years are effectively walled off rn. - Around 70% of $Nebius(NBIS.US) deals carry prepayments covering 50-60% of the capex, on take/pay terms. - $Coreweave(CRWV.US)'s $104B backlog is take/pay with ~21% of it recognized more than four years out. As we now know, $Coreweave(CRWV.US) signed an A100 contract running into 2029 (a SKU introduced in 2020) "at or above where it was years ago", and CEO told CNBC a batch of H100s coming off an expired contract was re-booked immediately at 95% of the original rate. $NVIDIA(NVDA.US)'s CFO also said that A100s shipped 6 years ago are still running at full utilization.And $NVIDIA(NVDA.US) is now willing to underwrite residual value itself, up to 25% per project in the new financing platforms. I don't think they'd be willing to backstop if they expect GPUs to be worthless in around 3 years or so.For the neoclouds themselves though, depreciation is the biggest cost driver e.g. Nebius D&A was larger than their adj EBITDA last Q. If economic life actually extends to a new base case scenario of 5-6 years rather than the old 3-4 years...Earnings power re-rates higher across $Coreweave(CRWV.US), $Nebius(NBIS.US) and even the hyperscalers as ultimate beneficiaries.GF Securities: CPO/NPO Update
NVIDIA Spectrum-X CPO Scale-OutProduction Status: Spectrum-X Co-Packaged Optics (CPO) switches for scale-out networking have officially entered mass production.Volume Forecasts:2026E: 15,000 units2027E: 100,000 unitsSupply Chain Ramp: TSMC has expanded CPO inspection equipment capacity—specifically adding Insertion 2/3 capacity—alongside supply chain progress in Fiber Array Units (FAU), shuffle boxes, and system assembly.Rubin Ultra Scale-Up Architecture & Optical EnginesTray Design Options: Rubin Ultra is slated to adopt a 9-18-9 tray design, though mechanical challenges associated with the 0.75U height profile could drive a potential reversion to a 10-9-8 layout.Design Impact: Architectural shifts in tray layout will have no impact on the optical engine (OE).NVIDIA Platform OE Shipments:2027E: 6 million units2028E: 19 million unitsTotal Industry-Wide OE Shipments:2027E: 11 million units2028E: 40 million unitsAmazon Trainium 4 & NPO DeploymentAWS Consumption Projections:2H27E: 5 million OE units2028E: 12 million OE units (primarily 6.4T specifications)Configuration & Packaging: Trainium 4 is projected to feature 3 distinct configurations, with two configurations expected to adopt Near-Package Optics (NPO).$Amazon(AMZN.US) $NVIDIA(NVDA.US)
Nvidia’s reduced OpenAI financing commitments and simultaneous $21B SpaceX stake highlight shifting AI‑capital priorities. Tech remains central as financing risks and strategic bets reshape focus.
Nvidia is scaling back a massive AI bet: the Wall Street Journal reported the chipmaker is cutting its financing guarantee for OpenAI's Ohio data center from $250B to under $120B, even as it disclosed...
Broadcom Inc. (AVGO) continues to solidify its role as a premier infrastructure backbone of the artificial intelligence boom. Currently trading around $392.99, the stock has generated notable traction following massive multi-year AI supply commitments from tech hyperscalers like Alphabet (Google), Meta, OpenAI, and Anthropic. Rather than directly fighting Nvidia in the GPU arena, Broadcom dominates the specialized market for custom AI accelerators (ASICs/XPUs) and high-performance networking technology, where management expects AI semiconductor revenue to exceed $100 billion by fiscal year 2027. This explosive demand is backed by an impressive backlog of new bookings exceeding $30 billion, giving the company stellar long-term revenue visibility into 2028.
Despite its stellar core fundamentals, investors should keep a close eye on emerging structural shifts and risk factors. While its custom chip business grew at a staggering 143% year-over-year pace to $10.8 billion last quarter, the stock has experienced recent volatility over its massive off-balance-sheet AI chip-financing vehicle. Analysts note that this specialized financing arm could carry up to $370 billion in senior debt by 2029. Though technically decentralized from Broadcom’s primary balance sheet, the company’s contingent multi-billion-dollar backstop on customer lease obligations creates a unique layer of customer concentration and credit risk if capital expenditures among tech giants start to cool.
Valuation-wise, AVGO remains an incredibly compelling narrative for growth-oriented portfolios. The stock trades at roughly 24.3 times forward earnings for fiscal 2027, making it remarkably cheaper than many of its high-flying semiconductor peers. For long-term investors looking to play the next phase of data center expansions without bearing extreme valuation risk, Broadcom presents a powerhouse setup heading into its upcoming September 2026 earnings report.
$NVIDIA(NVDA.US)
NVIDIA continues to be the dominant powerhouse driving the global AI revolution. Its robust data center revenue, cutting-edge GPU architecture, and strong ecosystem lock-in provide an unbeatable competitive moat.
My Strategy & Execution
I established this position with a focus on long-term growth rather than short-term price fluctuations. Holding through market swings allowed me to capture an impressive +40% gain so far.
While the momentum remains exceptionally strong, I will stay disciplined by monitoring upcoming earnings and macroeconomic shifts.
Karching 🥂
@Longbridge Singapore
@Captain's Treasure
$NVIDIA(NVDA.US)🔥 NVIDIA isn’t just selling chips anymore. It’s helping finance the AI revolution.
NVIDIA is heading into its Aug 26 earnings with investors watching whether AI spending is still accelerating. Expectations are already huge. (Investor’s Business Daily)
But here’s the interesting part: NVIDIA has teamed up with major Wall Street firms to help unlock up to $500 BILLION in capital for AI infrastructure.
That means more data centres → more AI compute → more demand for NVIDIA GPUs. 🚀
And Blackwell isn’t the end game. Vera Rubin is coming next, with NVIDIA saying the platform can dramatically reduce inference costs compared with Blackwell. (investor.nvidia.com)
The big question isn’t whether AI is growing.
It’s how long companies will keep spending billions to stay ahead.
$NVIDIA(NVDA.US) 🚀
AI infrastructure could be one of the biggest investment stories of this decade.
#NVIDIA #NVDA #AI #Semiconductors #StockMarket #Investing
$Intel(INTC.US)Intel: A Turnaround Story Worth Watching 🚀
Intel is one of the biggest names in semiconductors, but the company has faced major challenges in recent years. Now, its turnaround strategy and focus on advanced chip manufacturing could make the stock interesting again.
The key question is whether Intel can execute, regain technological leadership, and compete effectively against AMD, NVIDIA and TSMC.
If Intel succeeds, the upside could be significant. If execution continues to disappoint, investors could remain cautious.
Intel is definitely a stock to keep on the watchlist. 👀📈
#Intel #INTC #Stocks #Investing #Semiconductors #TechStocks
⚡ 𝐔𝐏𝐃𝐀𝐓𝐄: $HIVE Digital Tech(HIVE.US) HIVE Signs $350 Million Five-Year AI Cloud Deal for 2,016 $NVIDIA(NVDA.US) NVIDIA Blackwell Ultra GPUs
👉 𝗞𝗲𝘆 𝗛𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁𝘀:➤ 𝗛𝗜𝗩𝗘 signs five-year GPU cloud agreement worth approximately $𝟯𝟱𝟬 𝗺𝗶𝗹𝗹𝗶𝗼𝗻.➤ Deal adds approximately $𝟳𝟬 𝗺𝗶𝗹𝗹𝗶𝗼𝗻 in annualized revenue.➤ 𝗕𝗨𝗭𝗭 𝗛𝗣𝗖 total ARR rises to approximately $𝟭𝟴𝟬 𝗺𝗶𝗹𝗹𝗶𝗼𝗻.➤ Cluster will deploy 𝟮,𝟬𝟭𝟲 𝗡𝗩𝗜𝗗𝗜𝗔 𝗕𝗹𝗮𝗰𝗸𝘄𝗲𝗹𝗹 𝗨𝗹𝘁𝗿𝗮 𝗚𝗣𝗨𝘀.➤ Deployment uses 𝗚𝗕𝟯𝟬𝟬 𝗡𝗩𝗟𝟳𝟮 systems and Quantum-X800 InfiniBand networking.➤ AI infrastructure requires approximately $𝟭𝟴𝟱 𝗺𝗶𝗹𝗹𝗶𝗼𝗻 in capital expenditures.➤ Customer will provide an upfront deposit of approximately $𝟯𝟱 𝗺𝗶𝗹𝗹𝗶𝗼𝗻.➤ Cluster is expected online during 𝗤𝟰 𝟮𝟬𝟮𝟲 in British Columbia.➤ HIVE expects HPC/AI daily revenue around $𝟱𝟬𝟬,𝟬𝟬𝟬 after deployment.➤ HIVE targets $𝟮𝟬𝟬 𝗺𝗶𝗹𝗹𝗶𝗼𝗻 GPU cloud ARR by year-end.👉 𝗪𝗵𝘆 𝗧𝗵𝗶𝘀 𝗠𝗮𝘁𝘁𝗲𝗿𝘀:➤ Five-year contract provides HIVE with significant 𝗹𝗼𝗻𝗴-𝘁𝗲𝗿𝗺 𝗿𝗲𝘃𝗲𝗻𝘂𝗲 visibility.➤ Deal accelerates HIVE's diversification from 𝗰𝗿𝘆𝗽𝘁𝗼 𝗺𝗶𝗻𝗶𝗻𝗴 into AI infrastructure.➤ HIVE retains ownership of the GPUs, creating a 𝗹𝗼𝗻𝗴-𝗹𝗶𝘃𝗲𝗱 strategic asset.➤ Management sees potential to deploy over 𝟭𝟮𝟬,𝟬𝟬𝟬 𝗚𝗣𝗨𝘀 within two years.👉 𝗘𝘅𝗽𝗲𝗿𝘁 𝗦𝘁𝗮𝘁𝗲𝗺𝗲𝗻𝘁𝘀:𝗙𝗿𝗮𝗻𝗸 𝗛𝗼𝗹𝗺𝗲𝘀, Executive Chairman of HIVE Digital Technologies:"A year ago, our team was focused on rapidly scaling HIVE's ASIC computing power and Tier-I data center footprint by over 300%. Today, we are applying that same discipline and execution to scaling our GPU-based AI infrastructure.I am deeply pleased with the teamwork and execution of Aydin Kilic and Craig Tavares, together with the entire HIVE and BUZZ team, as we continue building our AI strategy. This new five-year agreement is another important milestone in that journey. With approximately 400 MW of capacity in Canada for Tier III data center development, we have the potential to bring over 120,000 GPUs online over the next 2 years.What excites me most is the synergy between our two businesses. Whether we are deploying ASIC chips for Tier-I data centers or GPU chips for AI, our core expertise remains the same: securing renewable energy, building high-performance data centers and converting that energy into valuable computing power. We believe this dual-engine strategy positions HIVE well for the next generation of digital infrastructure."𝗔𝘆𝗱𝗶𝗻 𝗞𝗶𝗹𝗶𝗰, President and Chief Executive Officer of HIVE Digital Technologies:"We are pleased to announce this agreement as we accelerate towards our year-end target of $200 million of ARR for our GPU cloud business. The five-year term, expected stable cash flows from an investment-grade enterprise customer, and upfront deposit of approximately 10% of total contract value supports an attractive economic and return profile for this deployment. Using proceeds from our June 2026 0% convertible bond, we are delivering on our promises to the market to bring this second, long-term large GPU cluster deal forward. We are also using debt financing to lever the purchase of the GPUs to provide a very strong internal rate of return."𝗖𝗿𝗮𝗶𝗴 𝗧𝗮𝘃𝗮𝗿𝗲𝘀, President and Chief Operating Officer of BUZZ High Performance Computing:"This contract demonstrates our ability to consistently execute on large-scale AI infrastructure deployments while securing long-term, high-quality customers. We're rapidly establishing BUZZ as one of Canada's leading sovereign AI cloud providers, combining best-in-class NVIDIA infrastructure with strategic partners such as Bell and VAST Data to deliver enterprise-grade AI compute at global scale.We're building far more than GPU clusters; we're helping build a sustainable AI future. This milestone reflects the speed and discipline with which our team continues to execute, while strengthening a world-class ecosystem of partners. Together, we're creating a sovereign AI cloud platform that gives Canadian and global organizations access to sustainable, high-performance infrastructure built for the AI era."
$NVIDIA(NVDA.US)
Nvidia is fast becoming not only the face of AI. She is also becoming the backbone. Everyone already recognises Nvidia as the ‘pioneer’ of AI, the one that woke the world up to AI.
Gone are the days when she was just a brand for graphic cards. (I remember those days when a wonderful gaming experience revolves around Nvidia graphics card and the ‘best games’ with amazing graphics need her to look at its best.)
Now she still does graphics cards of course but AI literally depends on her.
@Captain's Treasure
$NVIDIA(NVDA.US)
Its Nvidia time. Last Monday, Nvidia was going into the session trying to hold on to the 225 levels. It touched 224 before selling off to reach a low of 216.20 on Tuesday. Recovery began proper on Wednesday with a gap up and she held on to close the week at 225.
Today, premarket looks green with markets looking to open up including Nvidia.
236 is an important level. And Nvidia’s results are next week. A good result might provide the catalyst to propel her to new highs but having pampered the market with excellent results, let’s see how well she does this time.
@Captain's Treasure
$NVIDIA(NVDA.US) reducing the OpenAI guarantee is smart risk management. They still back AI, just not at $250B. The $21B SpaceX stake is a long-term moat play. SG +24.2% exports show the cycle isn’t dead.
Nvidia is scaling back a massive AI bet: the Wall Street Journal reported the chipmaker is cutting its financing guarantee for OpenAI's Ohio data center from $250B to under $120B, even as it disclosed...
Looks like money is rotating back to AI and specifically stocks that benefit from the buildout after they retreated in a healthy pullback.
Nvidia is scaling back a massive AI bet: the Wall Street Journal reported the chipmaker is cutting its financing guarantee for OpenAI's Ohio data center from $250B to under $120B, even as it disclosed...
Friday saw capital flowing into specific stocks related to the AI infrastructure. Overall, the markets are looking rather resilient. Nvidia’s earnings will be key as well as their forward guidance.
Nvidia is scaling back a massive AI bet: the Wall Street Journal reported the chipmaker is cutting its financing guarantee for OpenAI's Ohio data center from $250B to under $120B, even as it disclosed...
