Marvell Technology, Inc. MRVL shares rose Wednesday after the chipmaker unveiled a fresh capital return plan and highlighted its growing focus on data center and AI infrastructure. The company said its board approved a new $5 billion share repurchase authorization and disclosed an accelerated share repurchase agreement.Marvell’s board cleared a new buyback program totaling $5 billion. As of Aug. 2, 2025, the company had about $2.0 billion remaining under its prior repurchase authorization. Since then, it ha ...

Market Overview - The market is experiencing a rotation, with defensive sectors like utilities, real estate, and energy leading the way, while the majority of S&P 500 stocks remain in the green [2][3] - Current market conditions indicate a cooling off period, with investors awaiting new data and news flow, while still making higher highs and higher lows [4][5] - The S&P 500 has key levels to watch, with upside targets at 6690 and downside at 6630, indicating a potential for volatility [6][8] Alibaba Developments - Alibaba announced plans to exceed its previously set AI budget of $50 billion, focusing on building new data centers and expanding its cloud offerings [9][10] - The company is rolling out a new large language model, Quinn3 Max, which integrates with Nvidia's chipset, despite tensions with the Chinese government [11][12] - Alibaba's stock has seen a significant increase of approximately 9.7%, reflecting positive sentiment towards Chinese stocks with AI exposure [13][15] Lithium Americas Update - Lithium Americas' stock surged over 70% following reports of potential government involvement, with the company negotiating terms on a Department of Energy loan [19][20] - The negotiations may involve offering the U.S. government warrants for equity in exchange for better loan terms, indicating a strategic move to secure domestic lithium supplies [21][24] - The U.S. government aims to reduce reliance on China for lithium production, highlighting the importance of both mining and refining capabilities in the domestic supply chain [25][28]

Progress at our datacenter in Abilene. Fun to visit yesterday! https://t.co/W22ssjWstW ...

Data Center & Cloud Computing - Elon Musk claims to build a 10 GW data center before OpenAI [1] - NVIDIA reportedly plans to invest $100 billion to help OpenAI build a 10 GW data center [1] - OpenAI has placed a $300 billion cloud computing service order with Oracle over 5 years [1] - Oracle is procuring $40 billion worth of NVIDIA chips and leasing computing power to OpenAI [1] Companies Involved - NVIDIA (NVDA) is planning a significant investment in OpenAI's data center infrastructure [1] - Oracle (ORCL) is positioned to provide substantial cloud computing services and infrastructure to OpenAI [1]

Financial Data and Key Metrics Changes - The company reported a gross margin above 50%, the highest since mid-fiscal 2022, with operating margins also at their highest since November 2018 [32][33] - The company expects gross margins to improve further in the second quarter relative to the first quarter [33] Business Line Data and Key Metrics Changes - The NAND business is experiencing improved conditions, with expectations for tighter supply and increased demand driven by AI server deployments [6][7] - The DRAM segment is currently tight and expected to tighten further in 2026, contributing to improved pricing and margins [29][32] Market Data and Key Metrics Changes - The demand for high-capacity SSDs is increasing, with average capacities expected to escalate rapidly, particularly in AI servers [44] - The company anticipates significant growth in the data center segment, which has become a larger part of the total addressable market (TAM) and is driving overall profitability [54][55] Company Strategy and Development Direction - The company is focusing on enhancing its data center SSD business and has decided to exit the managed NAND market to improve overall ROI [9][71] - The strategy includes leveraging new product announcements and maintaining a strong competitive position in the data center SSD market [7][72] Management's Comments on Operating Environment and Future Outlook - Management expressed confidence in the demand for NAND and DRAM products, particularly due to the needs of hyperscalers for AI applications [6][7] - The company is optimistic about its ability to capture market share in HBM and expects to see higher share in HBM compared to previous years [13][18] Other Important Information - The company is investing significantly in DRAM construction and equipment, with CAPEX guidance increasing from approximately $13.8 billion in 2025 to about $18 billion in 2026, primarily for DRAM [11][19] - The company is ramping up production of one-gamma DRAM, which is expected to be the primary source of bit growth for fiscal 2026 [36] Q&A Session Summary Question: State of the NAND industry and pricing outlook - Management indicated that the bits down in the current quarter are noise based on segment mix and that demand from hyperscalers will drive NAND industry improvement [6][7] Question: HBM market share aspirations - The company expects to gain market share in HBM and is confident in its competitive positioning for HBM4, anticipating higher share compared to HBM3 [13][18] Question: CAPEX guidance and spending allocation - The majority of the increased CAPEX is directed towards DRAM construction and equipment, with minimal additional NAND spending [11][19] Question: DRAM revenue breakdown and margin contributions - Management clarified that while they do not provide specific margin comparisons, they expect tightening in the DRAM market to improve pricing and margins across the portfolio [29][32] Question: Long-term agreements with customers - There is interest in long-term agreements, but management is being cautious due to various market factors, including U.S. manufacturing and tariff implications [69][70] Question: Impact of exiting managed NAND on smartphone customers - Management confirmed that while customers may not be pleased with the exit from managed NAND, the strong relationship in DRAM remains intact [75]

Market Dynamics & Interconnectivity - NVIDIA Open AAI deal and Oracle's partnership with OpenAI create a circular dynamic, reminiscent of past tech announcements [1] - The interconnectedness of companies like Oracle, OpenAI, and Microsoft raises concerns about potential systemic risks [9][12] - The current AI build-out phase resembles past economic bubbles, where overspending and unmet demand led to economic slowdowns [8][9] Oracle's Financials & Strategy - Oracle's capital expenditure (capex) on data centers surpasses that of major hyperscalers [2] - Oracle's valuation is high, trading at approximately 40 times next year's earnings, which is historically high for the company [4] - Oracle has approximately $100 billion in debt and may experience negative free cash flow during the build-out phase [6] - Oracle might need external investment, potentially from Nvidia, to finance chip purchases [6] AI Investment & Revenue Potential - There is a lack of clarity regarding the revenue potential of current AI investments [13] - Enterprises are experimenting with AI, but the tangible benefits and productivity improvements are not yet evident [14] - A significant portion (50-60%) of capex for AI build-outs is attributed to Nvidia GPUs, indicating high demand [7] Financing & Debt Markets - Financing AI build-outs beyond major hyperscalers will require tapping into debt markets, including private credit [7][8] - Increased reliance on debt to fund capex cycles elevates financial risk [11] - Nvidia's partnership with Open AAI suggests a need to stimulate demand for its products, potentially signaling concerns about sustainable growth [12]

AI Infrastructure Expansion - Oracle, OpenAI, SoftBank are unveiling five more Stargate sites across Texas, New Mexico, Ohio, and an undisclosed location in the Midwest [1] - The initiative's revised spend is nearly 7 gigawatts and more than $400 billion investment over the next three years [2] - Total power demand from this and the Nvidia announcement reaches 17 gigawatts, representing an unprecedented infrastructure buildout in the US [9] Market Demand and Concerns - Oracle's CEO indicates broad-based demand across the industry, mitigating concerns about a market bubble due to committed demand [3] - OpenAI's CEO emphasizes that this scale of construction is necessary to keep up with AI's explosive growth, requiring significant infrastructure [4] - OpenAI anticipates needing much more compute to deliver on the horizon for AI [5] Powering AI Infrastructure - Nuclear power is being considered as a potential power source, with Sam Altman having supported a startup in this area [7] - Renewable energy and grid power in Texas may not be sufficient to meet the 7 gigawatts demand [8] Stock Market Reaction - Oracle's stock is expected to rise after hours following a 4% drop during the regular session [1] - Nvidia and Oracle both closed lower initially but are moving higher after hours, suggesting the deal is reigniting the AI trade [6]

This copy is for your personal, non-commercial use only. Distribution and use of this material are governed by our Subscriber Agreement and by copyright law. For non-personal use or to order multiple copies, please contact Dow Jones Reprints at 1-800-843-0008 or visit www.djreprints.com. ...

Investor Sentiment & Market Trends - Investor enthusiasm for data centers and AI is growing, with expectations that revenues will match or exceed expenditures in the coming years [1][2][3] - Investors are seeking diversification beyond the "Magnificent Seven" tech companies, looking for exposure to other semiconductor companies and data owners involved in the AI space [5][6] - AI is considered a long-term investment theme expected to drive markets for the next decade and beyond, despite potential short-term anxieties [15] ETF Performance & Strategy - One actively managed ETF has attracted over $5 billion this year, becoming the largest AI ETF in the United States, focusing on digital infrastructure, hardware producers, and data owners [2] - Actively managed funds can provide value by looking beyond consensus views and identifying emerging names in the AI ecosystem, globally [7][8] - The iShares innovation and tech actively managed ETF has seen accelerating fund flows, indicating a lack of significant anxiety among investors [14] Energy Sector & AI - There's increasing interest in various energy sources (nuclear, fossil fuels, renewables) to power the AI trade, viewing utilities through a forward-looking lens [11] - Investors are broadly considering the entire energy value chain, including necessary energy fuels, producers, and distribution methods [13] Company Strategy & Fundamentals - Leading technology companies with low costs of capital and access to dollars have a significant advantage in leveraging their strengths in the AI space [4] - Earnings direction is more important than current revenue for large language model developers, focusing on subscriber growth, model development, and corporate penetration [16][17]

Microsoft is testing out a possible solution to data center heat: sending fluid directly through tiny channels etched into chips https://t.co/kTreH1n3xE ...