AI Infographic - AI is compute and data intensive, and is poised to reshape our world. In this infographic we break down the phases of AI, associated storage requirements, and KIOXIA SSD solutions for AI. Check it out! #MakeItWithKIOXIA#

AI Infographic - AI is compute and data intensive, and is poised to reshape our world. In this infographic we break down the phases of AI, associated storage requirements, and KIOXIA SSD solutions for AI. Check it out! #MakeItWithKIOXIA#

Free O'Reilly Download: Designing Data Intensive Applications. In this practical guide, Martin Kleppmann helps you navigate the fast-changing landscape of processing & storing data for data-intensive apps

GLOBALFOUNDRIES CEO SAYS DATA-CENTER DEMAND IS A “VERY STRONG DRIVER” FOR GROWTH AS AI WORKLOADS BECOME INCREASINGLY DATA-INTENSIVE. CEO SAYS OPTICAL CONNECTIVITY IS NOW A “WHEN, NOT IF” TRANSITION AND GLOBALFOUNDRIES COULD 10X SILICON-PHOTONICS CAPACITY USING EXISTING US AND SINGAPORE FAB INFRASTRUCTURE.

AI agents are accelerating decisions and creating new data-intensive workloads, increasing the need for high-performance memory at the edge and scalable storage in the data center. At FMS 2026, Micron was recognized with three awards for innovations built to meet those demands:​ 🏆 Best of Show Award: 245TB​ Micron 6600 ION SSD 🌱 Sustainability Award: 245TB​ Micron 6600 ION SSD 💻 User Implementation Award: Micron LPCAMM2​ See the solutions driving the future of memory and storage from cloud to edge:

$840M+ USDT0 has moved onto @FlareNetworks - a chain purpose-built for data intensive use cases. When an AI agent uses Flare Time Series Oracle to make a decision, USDT0 is already there to settle the payment that follows. Data and dollars, on the same network.

Starlink’s reliable high-speed internet is onboard 30+ Goldenport vessels, enabling the maritime operator to run data intensive applications, modernize its operations, and provide crew with seamless connectivity while out at sea →

Codex stays focused and is surprisingly token efficient when given a well-specified goal along with well-aligned progress metrics. This was a somewhat complicated, greenfield project which required data acquisition from various sources, GPU-intensive data processing, and then custom software development for the primary deliverable. For these longer-range projects, I've found it useful to spend time with Codex completing at least one example, in full detail, end-to-end. Speaking aloud giving feedback (transcribed) provides useful context to help it understand decisions and rationale to assist it as it works to complete the full project in the coming days. "The completed autonomous goal used 66,968,831 tokens over 4 days, 1 hour, 11 minutes, and 5 seconds." @romainhuet @ajambrosino @thsottiaux

A TON OF THINGS HAPPENED IN THE STOCK MARKET TODAY. Here's a full recap: 1. Moderna $MRNA and Merck $MRK said their personalized mRNA cancer vaccine, intismeran, combined with Keytruda slowed melanoma recurrence and spread in a late-stage trial. The Phase 3 study was stopped at its first interim analysis after positive results, though detailed efficacy data have not yet been released. If the benefit holds, this would mark the first randomized Phase 3 validation of a personalized neoantigen cancer vaccine and could significantly expand the use case for mRNA technology beyond traditional vaccines. $MRNA surged 175% today, became the 2nd-best performing stock in the S&P 500 this year, and triggered an estimated $4.8B loss for short sellers. It was also the largest one-day percentage gain ever for an S&P 500 stock. 2. Crypto rallied sharply today as $BTC Bitcoin jumped roughly 8%, briefly touching about $70,000 for its biggest move since March, before settling near $68,500–$69,500. The move liquidated an estimated $1.4B–$1.7B of short positions and was driven by a mix of lower-yield pressure from Treasury buybacks, regulatory optimism, and today’s White House crypto meeting. $ETH Ethereum, $SOL Solana, and other major alts also traded higher, while crypto-linked stocks like Coinbase and Circle rallied. President Trump urged Congress to pass a “fair version” of the Clarity Act, said the U.S. should remain the leader in Bitcoin, crypto, prediction markets, and AI, and said he would listen to recommendations on whether the U.S. should accumulate more Bitcoin or crypto. Trump also said the CFTC is working to bring Hyperliquid into the U.S. in a fully compliant way, helping send HYPE higher. Overall, the tone was strongly pro-crypto, though the Clarity Act still faces political hurdles in the Senate. 3. Nebius $NBIS announced a proposed private offering of $4.5B of convertible senior notes. The offering includes $2.75B of notes due 2030 and $1.75B of notes due 2034, with an option for an additional $675M. Key terms, including interest rates and conversion prices, will be determined at pricing. Nebius was down 9.87% today. 4. U.S. national debt just crossed $40T, with the latest $10T added in roughly 4.5 years. The debt now equals about $116,481 per citizen and $360,794 per taxpayer, while the debt-to-GDP ratio sits around 122.7%. Federal spending is running at roughly $7.26T, compared to federal tax revenue of about $5.59T, leaving a federal budget deficit near $1.68T. Net interest on the debt is now running above $1.1T, making it one of the largest line items in the federal budget. 5. The U.S. Treasury announced it will double the size of long-term government debt buybacks after the rapid surge in Treasury yields. Repurchases of $2B will now increase to at least $4B, with the Treasury saying the move is meant to provide “liquidity support” for bonds maturing in 10 to 30 years as total U.S. debt nears $40T. The situation is like someone with a 30-year mortgage at 3.25% refinancing into a 1-year adjustable-rate loan at 5.75% because the market won’t absorb the longer-term debt cleanly. In other words, longer-duration, lower-cost debt is being replaced with shorter-term, higher-cost financing. Now apply that dynamic across trillions of dollars in U.S. government debt. That is the pressure Treasury is trying to manage as long-term yields keep rising. 6. Starlink is reapplying for India approval for its Gen 2 network, seeking authorization for nearly 30,000 LEO satellites operating at 340–615 km, including direct-to-device connectivity. India previously approved only Starlink’s 4,408-satellite Gen 1 network after rejecting the earlier Gen 2 application. The move comes as India is still finalizing D2D rules and satellite spectrum pricing, with Jio, Amazon Leo, and OneWeb also competing for the market, according to ET. 7. Goldman Sachs is preparing a roughly $1.15B junk bond to finance a Virginia data center leased to CoreWeave $CRWV. The deal is expected in September, with Goldman moving forward even as investors demand higher yields for AI infrastructure debt tied to CoreWeave. The financing comes after a separate $3.5B CoreWeave-backed deal recently priced at a 10% yield, underscoring how capital-intensive AI data center buildouts have become and how much credit markets are now focused on AI infrastructure risk. 8. The top 10 most active options today by contracts traded were $TSLA with 2.9M contracts, $NVDA with 2.6M contracts, $AAPL with 1.9M contracts, $AMZN with 1.0M contracts, $INTC with 932K contracts, $MU with 886K contracts, $MSTR with 789K contracts, $MSFT with 745K contracts, $SPCX with 694K contracts, and $META with 658K contracts. 9. Nvidia $NVDA has reportedly discussed investing in AI data supplier Mercor as part of a funding round that would value the startup at $20B, according to The Information. Mercor hires specialists across science, law, finance, and other fields to train and evaluate AI models, producing higher-quality data for complex tasks. Nvidia paid Mercor tens of millions of dollars last quarter and used its data to develop its two latest Nemotron models, with several Mercor employees now reportedly focused almost entirely on Nvidia-related work. The relationship is expanding as Nvidia invests more heavily in Nemotron, its open AI model family. Mercor has also worked with OpenAI, Google, and Anthropic, while annualized gross revenue reportedly reached $2B in June, double its level from the start of the year. 10. Webull $BULL reported Q2’26 revenue of $198.8M, beating estimates of $181M, up 51% YoY and 24% sequentially. Adjusted EPS came in at $0.12 versus $0.03 expected, up 140% YoY, while net income reached $24.4M and adjusted operating profit was $62.6M, representing a 31.5% margin. Trading-related revenue rose 66% YoY to $147.7M. Customer assets grew 79% YoY to $28.5B, registered users reached 28.2M, funded accounts hit 5.13M, equity notional volume rose 73% YoY to $279B, and options contract volume increased 68% YoY to 213M. Webull also launched Vega Analyst and announced the acquisition of Pi Securities in Thailand. 11. President Trump is reportedly set to reduce tariffs on automobiles imported from Canada from 25% to 15% as part of a broader trade deal, according to Bloomberg. The current tariff applies to the non-U.S. content of vehicles produced in Canada, and that same content provision is expected to apply under the new 15% rate. U.S. and Canadian negotiators have also discussed whether to expand the exemption to include additional content, though no final decision has been made. 12. U.S. margin debt fell by a record $85B in July to $1.42T, marking the largest monthly decline ever and the first monthly drop since March. For context, the prior record decline was $80B in January 2022, right as the bear market began. The pullback follows a massive $198B surge in margin debt across May and June, the largest two-month increase on record. Even after July’s historic decline, margin debt is still up $518B, or 58%, since the start of 2025. WALL STREET IS THE GREATEST SHOW ON EARTH.

Most of the smaller data-center companies won't survive over the long-term. - $RIOT - $BTDR - $MARA - $BRUN - $KEEL All of these will probably become irrelevant. Right now the market is still heavily supply constrained, so the tide lifts all boats. Here's an overview, which companies are actually positioned well, and which are not. 1. $CRWV - Currently biggest Neo Cloud with huge backlog - Lots of Know How for GPU Clusters 2. $NBIS - Full-Stack-AI-Cloud - Long-term contracts with $META and $MSFT 3. $IREN - Best miner Pivot, $MSFT and $NVDA contracts - Huge Renewables pipeline 4. $HUT - Huge long-term contracts with strong counterparties - Can survive as AI-Landlord 5. $APLD - Multiple big hyperscaler leases - less capital intensive than Neoclouds - Can survive as Power and Shell supplier 6. $WULF - Strong power access - Stays relevant as Power-Landlord, not as software player. 7. $CIFR - Strong HPC backlog - Low Time-To-Power Probably won't stay relevant: 8. $RIOT - Bad financing structure - Execution risk is high due to insufficient know how - Without cloud or custim silicon, Riot is a replaceable power and shell landlord 9. $BTDR - Cloud too small to keep up on depreciation and GPU-Generation swaps - Hybrid at mining and AI. No real focus on one model. - Lower customer quality than competition 10. $MARA - AI remains a secondary option behind mining - No high-density delivery or software layer - Only few running Take-or-Pay-AI-Leases 11. $KEEL - No real top tier contracts - Scale,power security and quality are lacking behind competitors - Currently still reliant on Mining 12. $BRUN - Very small Neocloud without power moat. - GPU-Capex needs to be contantly renewed - financing is a problem without a hyperscaler balance sheet. - High customer concentration and weak software layer. When supply overshoots demand, these will be the first to have problems selling their capacity. Any long-term investor needs to be aware of this. The buildout won't last forever. That's how capitalism works.