U.S. stocks were mixed Tuesday as semiconductor shares sold off sharply, sending the VanEck Semiconductor ETF (SMH) lower and weighing on the Nasdaq 100. Micron fell 9.5%, Corning dropped 8.9%, and NVIDIA declined 3.2%. The Nasdaq 100 slid 2.4% to 29,608, while the S&P 500 fell 0.9% and the Dow edged up 0.2%. Traders also questioned whether the AI trade has become overextended, with risk-off positioning intensifying ahead of Micron’s earnings.

Oklo and Centrus Energy signed a letter of intent for domestic high-assay low-enriched uranium (HALEU), with deliveries potentially beginning in 2029 to support up to five Aurora powerhouses over multiple years. The arrangement is framed as one of the early large-scale HALEU supply deals and may include prepayment structures from Oklo, directly addressing a key fuel constraint for commercialization. Oklo shares rose 1.64% to $59.36, but the stock remains in a bearish technical setup and faces resistance at $66.00, according to Benzinga Pro data.

Korn Ferry (NYSE: KFY) reported fourth-quarter results that topped expectations, with earnings of $1.40 per share versus forecasts of $1.38. Revenue came in at $768.256 million, above estimates of $743.340 million. The company operates in human resources and professional services, and the beat points to steady demand for hiring and organizational consulting. The declines in major stock indexes and commodities cited in the report reflect broader market performance and are not presented as directly caused by KFY’s earnings.

SpaceX’s market value has fallen by about $1 trillion, yet it successfully raised $20 billion through a bond sale and received investment-grade ratings from Moody’s, S&P Global and Fitch. The company is reported to hold roughly $100.8 billion in cash, while Starlink has grown to about 12 million subscribers. Figures circulating among investors indicate SpaceX could be generating approximately $2.32 billion per month from AI compute agreements with major AI companies. The episode highlights a split between equity investors focused on heavy spending and expectations of negative free cash flow, and bond investors emphasizing Starlink’s recurring cash flows and refinancing needs, with much of the proceeds expected to address obligations due in 2027.

A memory-themed ETF has risen 190% year to date, fueled by a surge in AI server demand for high-bandwidth memory (HBM). A potential Nasdaq IPO for SK Hynix and Micron’s fiscal-quarter results due Wednesday are shaping up as twin catalysts. Micron, one of the ETF’s largest holdings, is expected to shed light on HBM demand and pricing, which could bolster the case for memory ETFs as a core way to access AI infrastructure exposure.

Fervo Energy said it will work with NVIDIA and the Pacific Northwest National Laboratory to build EGS-Twin, a digital-twin platform that uses geothermal field data from Utah and Nevada to train AI models, with deployment targeted by 2029. The announcement was made the same day as the company released its first earnings report since going public. Fervo reported Q1 revenue of $61,000, well below the $480,000 analyst estimate, and said its net loss widened to $31.8 million. Even so, investors focused on the NVIDIA collaboration as a signal of technical validation and a potential boost to commercialization momentum.

埃森哲公布2025财年第三季度业绩：EPS为3.80美元，高于预期的3.69美元；营收187.2亿美元，略低于预期的187.5亿美元。公司下调2026财年收入指引至717.63亿–724.60亿美元，显著低于分析师预期的740.06亿美元，并调低Q4营收预期。受此影响，股价当日下跌2.7%。多家投行同步下调目标价。

Outdoor Holding Company（NASDAQ: POWW）公布2023年第四季度财报，每股亏损3美分，较分析师预期的亏损1美分更差；营收1389万美元，高于预期的1270万美元。受业绩不及预期影响，公司股价当日下跌超2%。文中提及的其他资产（如道指、纳指、黄金、原油等）均为市场整体表现或无关背景信息，未与POWW事件形成直接因果传导。

SK Hynix has overtaken Samsung Electronics in market capitalization, becoming South Korea’s most valuable company on the back of its leadership in high-bandwidth memory (HBM) chips. The company’s shares are up more than 327% year to date as heavy investment in AI computing infrastructure has boosted demand for advanced memory. As a key HBM supplier to NVIDIA, SK Hynix’s capacity and technology progress can affect the performance and delivery pace of GPUs used for AI training and inference.