After explosive gains in yesterday's session, Nvidia (NVDA -6.52%) stock is seeing big sell-offs in Thursday's session. The artificial intelligence (AI) hardware leader's share price was down 7.5% as of 1 p.m. ET. At the same point in today's trading, the S&P 500 was down 4.5%, and the Nasdaq Composite was down 5.5%.The stock market saw nearly unprecedented single-day gains yesterday after President Donald Trump announced that goods from all countries except for China would have an import tax rate of 10% an ...

Nvidia Corporation’s (NASDAQ: NVDA ) reputation as market’s AI darling has undoubtedly been in deterioration as of late. The stock’s declined by more than 30% this year, with investor confidence increasingly blighted by a concoction of risks to Nvidia’s demandAnalyst’s Disclosure: I/we have a beneficial long position in the shares of NVDA either through stock ownership, options, or other derivatives. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other ...

Core Viewpoint - Nvidia has been a significant beneficiary of the generative AI boom, with its stock price increasing dramatically since the launch of ChatGPT in November 2022, peaking at $149.43 per share by January 2025 [1] Group 1: Stock Performance and Historical Context - Nvidia's stock has experienced a 35% decline recently due to trade tensions and concerns over AI infrastructure spending sustainability [2] - Historically, Nvidia shares have rebounded after similar declines, averaging a 305% return in the two years following drawdowns exceeding 35% [4] - Over the past decade, Nvidia's stock price increased by 17,400%, despite facing three significant drawdowns of over 35% [3][4] Group 2: Market Position and Demand - Nvidia holds a dominant market position with approximately 98% market share in data center GPUs and over 85% in AI accelerators [6] - Concerns have arisen regarding competition from DeepSeek, a Chinese AI startup, which may impact demand for Nvidia's GPUs [7] - Despite potential competition, analysts expect overall demand for Nvidia GPUs to increase as AI becomes more accessible and costs decrease [9] Group 3: Future Growth Opportunities - Nvidia is well-positioned to benefit from various applications beyond generative AI, including autonomous vehicles and robotics, which will drive demand for its products [9][10] - The AI accelerator market is projected to grow at an annual rate of 29%, suggesting durable revenue growth for Nvidia [13] - Wall Street anticipates Nvidia's earnings to grow at 38% annually through fiscal 2027, indicating that the current valuation of 32 times earnings is attractive [14] Group 4: Competitive Landscape - While Nvidia may face increased competition from ASICs developed by Broadcom and Marvell, the threat is not as significant as perceived [11] - Nvidia's GPUs are supported by a comprehensive suite of software development tools, which provides a competitive edge over ASICs that require companies to develop applications from scratch [12] - Despite potential market share loss, Nvidia's GPUs are expected to account for over 80% of AI accelerator sales by 2030 [13] Group 5: Investment Outlook - Nvidia shares are currently down 35% from their peak, but the company has a history of rebounding from larger losses [15] - The long-term growth potential and current valuation suggest that Nvidia stock is a worthwhile investment for patient investors [15]

Core Insights - Google introduced its seventh-generation TPU, Ironwood, at the Google Cloud Next 25 conference, marking it as the most powerful TPU to date designed for large-scale AI reasoning models [1][2] TPU Overview - TPU (Tensor Processing Unit) is a specialized AI chip designed to accelerate deep learning tasks, first introduced by Google in 2015, with the first generation released in 2016 [2] - Ironwood represents a shift in AI infrastructure from reactive models providing real-time information to proactive models generating insights and interpretations [2] Technical Specifications - Ironwood can support a maximum cluster of 9,216 liquid-cooled chips, achieving a peak performance of 42.5 ExaFlops, equating to 42.5 quintillion operations per second [2] - The chip supports FP8 computation, with a performance of 4,614 TFlops, slightly surpassing NVIDIA's B200 at 4,500 TFlops, and has a memory bandwidth of 7.2 TBps, which is lower than B200's 8 TBps [3] - Ironwood features the third-generation SparseCore accelerator, designed to enhance financial and scientific computations, initially aimed at accelerating recommendation models [3] Comparative Analysis - Comparison of TPU generations shows significant advancements: - Pod Size: Ironwood (9,216 chips) vs. TPU v4 (4,896 chips) and TPU v5p (8,960 chips) - HBM Bandwidth: Ironwood (192 GB) vs. TPU v4 (32 GB) and TPU v5p (95 GB) - Capacity: Ironwood (7.4 TBs HBM) vs. TPU v4 (1.2 TBs HBM) and TPU v5p (2.8 TBs HBM) - Peak performance per chip: Ironwood (4,614 TFLOPS) vs. TPU v4 (275 TFLOPS) and TPU v5p (459 TFLOPS) [4] - Ironwood's performance per watt is double that of the previous generation TPU, Trillium, and its chip capacity is six times larger, allowing for handling of larger models and datasets [4] Future Plans - Google plans to integrate TPU v7 into its cloud AI supercomputing services, which will include recommendation algorithms, Gemini models, and AlphaFold [4] - OpenAI's co-founder Ilya Stutskever's AI startup, Safe Superintelligence, is utilizing Google Cloud's TPU chips for its AI research [5]

Core Viewpoint - The U.S. government has decided to suspend further restrictions on NVIDIA's H20 AI chip exports to China following a dinner attended by NVIDIA's CEO Jensen Huang at Mar-a-Lago [1] Group 1: Government Policy Changes - The White House has altered its plans to limit NVIDIA's exports of the H20 chip to China after the recent engagement with the company's CEO [1] - The decision indicates a potential shift in the U.S. government's approach towards regulating AI technology exports [1] Group 2: Industry Implications - NVIDIA, as a leading player in the AI chip market, may benefit from the relaxed export restrictions, potentially enhancing its market position in China [1] - The suspension of restrictions could lead to increased sales and partnerships for NVIDIA in the Chinese market, which is significant for its growth strategy [1]

Core Viewpoint - The Trump administration has temporarily halted plans to expand export restrictions on Nvidia's H20 AI chips to China, following Nvidia CEO Jensen Huang's commitment to invest in U.S. AI data centers [1][3]. Group 1: Nvidia's Position and Actions - Nvidia's CEO Jensen Huang attended a dinner at Trump's Mar-a-Lago estate, where discussions about the H20 chip and potential investments in U.S. AI infrastructure took place [1][3]. - Huang indicated that Nvidia may invest approximately $500 billion in U.S.-made chips over the next four years, highlighting the company's commitment to domestic production [2]. Group 2: Government and Legislative Context - The H20 chip is considered the most advanced AI chip that U.S. companies can legally sell to China, amidst ongoing pressures from U.S. lawmakers for stricter export controls on AI technologies [3][5]. - The U.S. House of Representatives' "China Strategic Competition Special Committee" expressed disappointment over the delay in imposing restrictions on the H20 chip, emphasizing the urgency of limiting exports to prevent competitors from gaining an advantage [5]. Group 3: Financial Performance of Nvidia - Nvidia reported a record revenue of approximately $130.5 billion for the fiscal year ending January 26, 2025, representing a 114% year-over-year increase, with a net profit of about $72.9 billion, up 145% [5]. - The company's revenue from China reached $17.1 billion, marking its highest ever, and 53% of Nvidia's revenue for the fiscal year came from outside the U.S. [5].

21世纪经济报道新质生产力研究院综合报道 早上好，新的一天又开始了。在过去的24小时内，科技行业发生了哪些有意思的事情？来跟21tech一起 看看吧。 【巨头风向标】 英伟达开源一款推理模型。4月9日，英伟达开源了一款名为Llama-3.1-Nemotron-Ultra-253B-v1的推理模 型。该模型基于Meta早期的Llama-3.1-405B-Instruct模型开发微调，有2530亿个参数。英伟达新模型的 性能逼近拥有6710亿参数的 DeepSeek R1，但只用了不到一半的参数量；在多项基准测试中，甚至击败 了两款刚发布的Llama 4模型。并且，Llama-3.1-Nemotron-Ultra-253B推理吞吐量也比DeepSeek R1 671B 高4倍。不过，在MATH500和Arena Hard基准测试中，DeepSeek R1仍略占优势。 谷歌云发布新一代TPU、大模型。在谷歌云下一步（Google Cloud Next '25）大会上，谷歌发布了第七 代张量处理单元（TPU）——Ironwood。与之前同时兼顾训练和推理的TPU不同，Ironwood是谷歌首款 专为推理工作负载设计的 ...

Nvidia CEO Jensen Huang appears to have struck a deal with the Trump administration to avoid export restrictions on the company’s H20 AI chips.The H20, the most advanced Nvidia-produced AI chip that can still be exported from the U.S. to China, was reportedly spared thanks to a promise from Huang to invest in new AI data centers in the U.S. According to NPR, Huang made the proposal during a dinner at Trump’s Mar-a-Lago resort sometime last week. Nvidia declined to comment.Many in the semiconductor industr ...

I'm a full time value investor and writer who enjoys using classical value ratios to pick my portfolio. Long-term focused on low P/B, P/FCF, Owner Earnings discounting, PEG ratios, the Graham Number and an occasional net-net hunter. I also believe tracking earnings growth versus price appreciation is an essential element to any quality evaluation.I also advocate self-indexing primarily using the Dow Jones Industrial Average as my index of choice combined with Joel Greenblatt's Magic Formula. This can help w ...

金十数据4月10日讯，美股"七巨头"持续飙升，特斯拉(TSLA.O)涨近18%，英伟达(NVDA.O)涨超15%， Meta Platforms(META.O)涨近12%，亚马逊(AMZN.O)、苹果(AAPL.O)涨近10%，微软(MSFT.O)涨超 8%，Alphabet(GOOG.O) 涨近8%。 美股"七巨头"持续飙升 ...