The Hottest IPO of 2026 Will Shock You

以下是内容的中文翻译： **《Motley Fool Hidden Gems Investing》最新一期节目以讨论英伟达（Nvidia）在蓬勃发展的人工智能（AI）领域的大手笔交易拉开序幕，随后转向一家中国存储芯片制造商首次公开募股（IPO）的影响，并以展望并购格局的变化作结。** **英伟达大规模的人工智能投资** 节目主持人Travis Hoyam、Lou Whiteman和Tyler Crow首先讨论了黄仁勋（Jensen Huang，在文字记录中被称为Jensen Wong）和英伟达正在进行的可能“大规模”交易的消息。《华尔街日报》报道称，英伟达可能为OpenAI一个价值2500亿至5000亿美元的项目提供资金担保，涵盖设施建设和芯片供应，这还不包括其现有的股权投资。英伟达还在与其他AI初创公司，如Thinking Machines和Safe Superintelligence，进行交易。 Tyler Crow强调了这些数字的巨大规模，质疑它们是否“只是凭空捏造”，而Travis Hoyam将其与礼来公司（Eli Lilly）通过高估值和现金围绕其业务构建“护城河”的策略进行了比较。Lou Whiteman指出英伟达强大的资产负债表，这缓解了对公司失败的担忧，但他警告说，成功取决于人工智能能否实现其崇高的收入预测。这些交易的“循环会计”或“供应商融资”方面，让人想起1990年代的科技泡沫，引发了担忧。然而，一个关键的区别是：90年代出现了光纤产能过剩，而今天的数据中心空置率很低，这表明当前投资是由真实需求驱动的。尽管如此，债务、特殊目的载体（SPV）和义务担保的参与仍被视为潜在的泡沫指标，甲骨文（Oracle）的股票表现（自2023年末其与OpenAI达成的3000亿美元交易以来下跌了63%）被引作一个警示性例子，说明尽管有巨额“欠条”，市场信心仍可能减弱。 **长鑫存储（CXMT）的IPO与存储市场动态** 对话随后转向了长鑫存储（CXMT），一家中国存储制造商，该公司最近在中国上市，股价飙升近500%。长鑫存储目前是第四大存储制造商，目标是迅速扩大其DRAM（动态随机存取存储器）产能，可能成为第三大制造商。这一发展为由美光（Micron）、SK海力士（SK Hynix）和三星（Samsung）长期主导的市场引入了新的动态。 Lou Whiteman指出，虽然这对长鑫存储有利，但这种扩张可能导致商品化加剧，并影响现有厂商的供应和定价。主持人讨论了地缘政治影响，五角大楼将长鑫存储列为与中国军方有关联的企业，这引发了关于苹果（Apple）等美国公司能否从他们那里采购芯片的问题。Tyler Crow区分了长鑫存储目前专注于低带宽、商品化DRAM与AI数据中心使用的高带宽存储器（HBM）。然而，长鑫存储在2-3年内开发HBM的既定目标，可能达到SK海力士过去的水平，构成未来威胁。产能过剩的风险并非长鑫存储独有；美光公司本身也计划在未来十年内投入2500亿美元进行产能扩张。虽然DRAM供应的涌入可能降低消费电子产品的价格并填补市场空白，惠及消费者，但也带来了“意想不到的后果”的风险，如果现有厂商为了追逐数据中心市场而忽视核心客户。 **并购狂潮与并购格局的变迁** 最后，主持人探讨了当前美国并购（M&A）环境，注意到政府对并购的态度似乎变得更加宽松。例子包括华纳兄弟探索（Warner Bros. Discovery）与派拉蒙（Paramount）潜在的合并、铁路公司整合谈判（诺福克南方铁路（Norfolk Southern）/联合太平洋铁路（Union Pacific）），甚至联合航空（United Airlines）和达美航空（Delta Airlines）首席执行官之间关于可能合并的讨论。 Tyler Crow认为公司正在抓住机会，将其视为“趁势而为”的局面，尤其是考虑到前一届政府对像Spirit和JetBlue这样的小型交易都曾进行阻挠。然而，Lou Whiteman表示怀疑，指出首席执行官们可能只是根据新闻报道“异想天开”。他强调，尽管最初获得绿灯，一些交易仍面临显著延迟（派拉蒙/华纳兄弟探索的交易已搁置一年）以及联邦和州监管机构的持续审查。即使是科技领域常见的较小的“附加”收购（tack-on acquisitions）也面临挑战（例如亚马逊/iRobot）。结论是，尽管并购环境可能比几年前更宽松，但由于持续的不确定性和多层监管，它不一定是“新的黄金时代”，这表明“意想不到的后果”可能是今年余下时间的主题。

The latest episode of Motley Fool Hidden Gems Investing opened with a discussion on Nvidia's aggressive deal-making in the burgeoning AI sector, moved to the impact of a Chinese memory chip manufacturer's IPO, and concluded with a look at the shifting landscape of mergers and acquisitions. **Nvidia's Massive AI Investments** The hosts, Travis Hoyam, Lou Whiteman, and Tyler Crow, first tackled news that Jensen Huang (referred to as Jensen Wong in the transcript) and Nvidia are making potentially "massive deals." A Wall Street Journal report indicated Nvidia might backstop a $250 billion to $500 billion project for OpenAI, covering facility build-out and chips, in addition to an existing equity stake. Nvidia is also making deals with other AI startups like Thinking Machines and Safe Superintelligence. Tyler Crow highlighted the sheer scale of these numbers, questioning if they're "just made up," while Travis Hoyam drew parallels to Eli Lilly's strategy of using high valuation and cash to build a moat around its business. Lou Whiteman noted Nvidia's strong balance sheet, alleviating concerns about the company's failure, but cautioned that the success hinges on AI meeting its lofty revenue projections. The "circular accounting" or "vendor financing" aspect of these deals, reminiscent of the 1990s tech bubble, raised concerns. However, a key distinction was made: while the 90s saw fiber optic overcapacity, today's data center vacancy rates are low, suggesting a real demand driving current investments. Still, the involvement of debt, special purpose vehicles (SPVs), and guarantees of obligations were flagged as potential bubble indicators, with Oracle's stock performance (down 63% since its $300 billion OpenAI deal in late 2023) cited as a cautionary example of how market confidence can wane despite large "IOUs." **The CXMT IPO and Memory Market Dynamics** The conversation then shifted to CXMT (ChangXin Memory Technologies), a Chinese memory maker that recently had an IPO in China, seeing its stock surge nearly 500%. CXMT, currently the fourth-largest memory maker, aims to rapidly expand its DRAM capacity, potentially becoming number three. This development introduces a new dynamic to a market historically dominated by Micron, SK Hynix, and Samsung. Lou Whiteman noted that while beneficial for CXMT, this expansion could lead to increased commoditization and impact supply/pricing for existing players. The hosts discussed the geopolitical implications, with the Pentagon classifying CXMT as a Chinese military operation, raising questions about U.S. companies like Apple's ability to procure chips from them. Tyler Crow differentiated CXMT's current focus on lower-bandwidth, commoditized DRAM from the high-bandwidth memory (HBM) used in AI data centers. However, CXMT's stated goal to develop HBM within 2-3 years, potentially matching SK Hynix's past capabilities, poses a future threat. The risk of overcapacity is not unique to CXMT; Micron itself plans to spend $250 billion over the next decade on capacity expansion. While an influx of DRAM supply could lower prices for consumer electronics and fill market gaps, benefiting consumers, it also carries the risk of "unintended consequences" if established players neglect core customers by chasing the data center market. **Merger Mania and Shifting M&A Landscape** Finally, the hosts explored the current U.S. merger and acquisition (M&A) environment, noting a perceived shift towards a more lenient approach from the administration. Examples included the potential Warner Bros. Discovery/Paramount merger, railroad consolidation talks (Norfolk Southern/Union Pacific), and even discussions between United Airlines and Delta Airlines CEOs about a possible merger. Tyler Crow suggested that companies are seizing the opportunity, viewing it as a "getting while the getting's good" situation, especially given past administration pushback on smaller deals like Spirit and JetBlue. However, Lou Whiteman introduced skepticism, pointing out that CEOs might be "daydreaming" based on press clippings. He highlighted that despite initial green lights, some deals face significant delays (Paramount/WBD is on hold for a year) and ongoing scrutiny from federal and state regulators. Even smaller "tack-on" acquisitions, common in tech, have faced challenges (e.g., Amazon/iRobot). The conclusion was that while the M&A environment might be more permissive than a few years ago, it's not necessarily a "new golden age" due to lingering uncertainties and multi-level oversight, suggesting that "unintended consequences" might be the defining theme for the remainder of the year.