NVIDIA accelerated computing starts with DGX, the world's AI supercomputer, the engine behind the large language model breakthrough. IHand delivered the world's first DGX to open AI. Since then, half of the Fortune 100 companies have installed DGX AI supercomputers. DGX has become the essential instrument of AI. The GPU of DGX is eight H100 modules.H100 has a transformer engine designed to process models like the amazing chat GPT which stands for generative pre-trained transformers. The eight H100 modules a ...

本文来自微信公众号：凤凰网科技 （ID：ifeng_tech），作者：姜凡，编辑：董雨晴，原文标题：《寒武纪捅破了天？半年报炸裂、净利润超10亿》，题 图来自：AI生成 8月26日，寒武纪（688256.SH）发布的半年报，直接在资本市场投下了一枚重磅炸弹。2025年上半年，公司实现营业收入28.81亿元，同比增长 4347.82%；归属于母公司股东的净利润10.38亿元，而去年同期还是亏损5.3亿元，如今一举扭亏为盈，基本每股收益也从-1.27元翻红至2.5元。 | | | | 单位: 元 中神:人民币 | | | --- | --- | --- | --- | --- | | | 本报告期末 | 上年度末 | 本报告期末比上年度末 | | | | | | 增减(%) | | | 总资产 | 8, 420, 117, 458. 26 | 6, 717, 812, 509. 70 | | 25. 34 | | 归属于上市公司股 | 6,755, 397, 722. 83 | 5, 422, 658, 659. 68 | | 24. 58 | | 东的净资产 | | | | | | | 本报告期 | 上年同期 ...

马斯克宣布了一个疯狂的计划，将在5年内实现5000万张H100的算力，这是什么概念？这将为人类带来怎样的影响？ASI能否在勇敢者的孤注 一掷下现身？ 世界首富马斯克，这次宣布决定All in AI了。 5年内实现5000万张H100的算力。 要知道，他已经有了全世界最强的Colossus超算集群，AI算力等价于约20万张H100。 他究竟想用这么多GPU做些什么呢？ 十万亿元能创造出怎样的奇迹 目前，每张H100的批发价高达2万美元。 5000万张H100，光是GPU，成本就将高达1万亿美元。 要搭建目前的最先进的超算集群，目前GPU成本只占约50%。 也就是说，最终的成本将超过2万亿美元（逾14万亿元人民币）。 2万亿美元是什么概念？ 美国去年的军费总支出约9970亿美元，而这已经占到了全球军费支出的37%。 这意味着，AI已经成为与传统的军备竞赛分庭抗礼的全新关键领域。 马斯克的身价约4000亿美元。 特斯拉的市值约1.1万亿美元。 加上SpaceX、X和xAI，马斯克旗下的公司市值约1.6万亿美元。 一旦摩尔定律在未来5年不能在GPU上有效，成本将无法产生指数下降。 马斯克是在拉上自己和全体股东的全部身 ...

新智元报道 编辑：艾伦 桃子 【新智元导读】马斯克宣布了一个疯狂的计划，将在5年内实现5000万张H100的算力，这是什么概念？这将为人类带来怎样的影响？ASI能否在勇敢者的 孤注一掷下现身？ 世界首富马斯克，这次宣布决定All in AI了。 5年内实现5000万张H100的算力。 要知道，他已经有了全世界最强的Colossus超算集群，AI算力等价于约20万张H100。 他究竟想用这么多GPU做些什么呢？ 十万亿元能创造出怎样的奇迹 目前，每张H100的批发价高达2万美元。 5000万张H100，光是GPU，成本就将高达1万亿美元。 要搭建目前的最先进的超算集群，目前GPU成本只占约50%。 也就是说，最终的成本将超过2万亿美元（逾14万亿元人民币）。 2万亿美元是什么概念？ 美国去年的军费总支出约9970亿美元，而这已经占到了全球军费支出的37%。 这意味着，AI已经成为与传统的军备竞赛分庭抗礼的全新关键领域。 马斯克的身价约4000亿美元。 特斯拉的市值约1.1万亿美元。 加上SpaceX、X和xAI，马斯克旗下的公司市值约1.6万亿美元。 一旦摩尔定律在未来5年不能在GPU上有效，成本将无法产生指数 ...

Salt Lake City, Aug. 26, 2025 (GLOBE NEWSWIRE) -- BluSky AI Inc. (OTCID: BSAI) (“BluSky AI” or the “Company”), Headquartered in Salt Lake City, Utah, BluSky AI Inc. is a Neocloud purpose-built for artificial intelligence through rapidly deployable SkyMod data centers. SkyMods are next-generation, scalable AI Factories. As a provider of GPU-as-a-Service, today announced the signing of a Letter of Intent (LOI) with Lilac, a next-generation GPU marketplace platform. This agreement marks the beginning of a stra ...

今天，我们就会从这些维度，带你拨开数据的表面，看到背后真正的逻辑。本次分析，将是全网独一无二的深度解读。如果你想领先市场一步，千万不要 错过任何一个细节。 英伟达未来三大催化剂 英伟达如今的市值已经飙到 4.34 万亿美元（截止上周五），比日本一整年的GDP还要高。 英伟达之所以这么牛绝不仅仅因为GPU卖得好，那只是表层的故事。真正让资本市场疯狂的是，它数次财报里透露出来的对未来AI时代垄断地位的巨大 预期。 我们美股投资网是唯一敢在英伟达财报前，明确亮明自己观点的频道！从2024年2月19日至今，我们已经连续做了六次英伟达财报前瞻，其中 5次全部正 确。 这一次我们也参考了大量的资料，力求用数据说话！ 我们这次对英伟达财报判断是：大概率超预期。但重点在于，英伟达的股价已经提前消化了大量利好，想要在这个位置来一波大涨，难度比之前可要大得 多。换句话说，涨没问题，但不期望会暴涨。 曾成功投资过脸书推特Robinhood和Coinbase的美国顶级的风险投资（a16z）表示过，英伟达的优势，并非只是GPU性能，还有其围绕芯片，在网络、内 存、供应链及产业生态上建立的护城河。 问题是，资本市场的逻辑残酷——没有永远 ...

CoreWeave, Inc. CRWV stock is at a crossroads, with the NVIDIA Corp. NVDA earnings report preparing to test investor confidence and the company's long-term trajectory. CRWV stock is down 20% this month. Check out the chart here. CoreWeave's business model is deeply intertwined with Nvidia's ecosystem as it provides high-performance GPU cloud services, largely powered by Nvidia's H100 and A100 chips. This dependence means Nvidia's financial performance, guidance and commentary on AI demand has the potential ...

Core Viewpoint - NVIDIA Corporation (NVDA) is expected to report strong earnings for the second quarter of fiscal 2026, with projected revenues of $45 billion, reflecting a 53.2% year-over-year increase, although slightly below the consensus estimate of $46.03 billion [1][8]. Revenue Projections - The anticipated revenue for NVIDIA's Data Center business is $40.19 billion, indicating a robust year-over-year growth of 53% driven by demand for AI and cloud chips [7][8]. - The Gaming segment is projected to generate $3.81 billion in revenue, representing a 32.4% increase from the previous year [9]. - The Professional Visualization segment is estimated to achieve revenues of $529.1 million, reflecting a 16.5% year-over-year growth [10]. - The Automotive segment is expected to report revenues of $591.6 million, indicating a significant year-over-year growth of 67.7% [11]. Earnings Estimates - The Zacks Consensus Estimate for quarterly earnings has increased to $1.00 per share, suggesting a year-over-year growth of 47.1% from 68 cents per share [2]. - The Earnings ESP for NVIDIA is +3.14%, indicating a strong likelihood of an earnings beat this quarter [5]. Market Performance - NVIDIA's stock has increased by 35.3% over the past year, outperforming the Zacks Computer and Technology industry's growth of 18.7% [12]. - The company trades at a forward P/E of 34.78X, which is higher than the sector average of 27.24X, indicating a premium valuation [14]. Industry Trends - The global generative AI market is projected to reach $967.6 billion by 2032, with a CAGR of 39.6% from 2024 to 2032, driving demand for NVIDIA's AI chips [20]. - NVIDIA's dominance in the generative AI chip market positions it favorably for substantial revenue growth as industries modernize their workflows [21]. Investment Considerations - NVIDIA's strong product portfolio and leadership in AI and data centers present a compelling investment opportunity, although its high valuation may lead to short-term volatility [22].

Core Insights - The article discusses the cost comparison between H100 and GB200 NVL72 servers, highlighting that the total upfront capital cost for GB200 NVL72 is approximately 1.6 to 1.7 times that of H100 per GPU [2][3] - It emphasizes that the operational costs of GB200 NVL72 are not significantly higher than H100, primarily due to the higher power consumption of GB200 NVL72 [4][5] - The total cost of ownership (TCO) for GB200 NVL72 is about 1.6 times higher than that of H100, indicating that GB200 NVL72 needs to be at least 1.6 times faster than H100 to be competitive in terms of performance/TCO [4][5] Cost Analysis - The price of H100 servers has decreased to around $190,000, while the total capital cost for a typical hyperscaler server setup can reach $250,866 [2][3] - For GB200 NVL72, the upfront capital cost per server is approximately $3,916,824, which includes additional costs for networking, storage, and other components [3] - The capital cost per GPU for H100 is $31,358, while for GB200 NVL72, it is $54,400, reflecting a significant difference in initial investment [3] Operational Costs - The operational cost per GPU per month for H100 is $249, while for GB200 NVL72, it is $359, indicating a smaller margin in operational expenses [4][5] - The electricity cost remains constant at $0.0870 per kWh across both systems, with a utilization rate of 80% and a Power Usage Effectiveness (PUE) of 1.35 [4][5] Recommendations for Nvidia - The article suggests that Nvidia should enhance its benchmarking efforts and increase transparency to benefit the machine learning community [6][7] - It recommends expanding benchmarking beyond NeMo-MegatronLM to include native PyTorch, as many users prefer this framework [8][9] - Nvidia is advised to improve diagnostic and debugging tools for the GB200 NVL72 backplane to enhance reliability and performance [9][10] Benchmarking Insights - The performance of training models like GPT-3 175B using H100 has shown improvements in throughput and efficiency over time, with significant gains attributed to software optimizations [11][12] - The article highlights the importance of scaling in training large models, noting that weak scaling can lead to performance drops as the number of GPUs increases [15][17] - It provides detailed performance metrics for various configurations, illustrating the relationship between GPU count and training efficiency [18][21]

Summary of Conference Call on Scale Up Network Growth from NVIDIA's Perspective Industry Overview - The discussion revolves around the **Scale Up network** in the context of **NVIDIA** and its implications for the broader **computing power** industry, particularly in AI and parallel computing applications [1][5][9]. Core Insights and Arguments - **Scaling Law**: The concept of a "Scaling Law" in networks is proposed, emphasizing the need for larger cross-cabinet connections rather than just existing ASIC and cabinet solutions [1][5]. - **NVIDIA's Strategy**: NVIDIA aims to address hardware memory wall issues and parallel computing demands by increasing **Nvlink bandwidth** and expanding the **Up scale** from H100 to GH200, although initial adoption was low due to high costs and insufficient inference demand [6][8]. - **Memory Wall**: The memory wall refers to the disparity between the rapid growth of model parameters and computing power compared to memory speed, necessitating more HBM interconnect support for model inference and GPU operations [1][10]. - **Performance Metrics**: The GB200 card shows significant performance differences compared to B200, with a threefold performance gap at 10 TPS, which increases to sevenfold at 20 TPS, highlighting the advantages of Scale Up networks under increased communication pressure [4][14][15]. - **Future Demand**: As Scale Up demand becomes more apparent, segments such as **fiber optics**, **AEC**, and **switches** are expected to benefit significantly, driving market growth [9][28]. Additional Important Points - **Parallel Computing**: The evolution of computing paradigms is shifting towards GPU-based parallel computing, which includes various forms such as data parallelism and tensor parallelism, each with different communication frequency and data size requirements [11][12]. - **Network Expansion Needs**: The need for a second-layer network connection between cabinets is emphasized, with recommendations for using fiber optics and AEC to facilitate this expansion [4][23][24]. - **Market Trends**: The overall network connection growth rate is anticipated to outpace chip demand growth, benefiting the optical module and switch industries significantly [28][30]. - **Misconceptions in Market Understanding**: There is a prevalent misconception that Scale Up networks are limited to cabinet-level solutions, whereas they actually require larger networks composed of multiple cabinets to meet user TPS demands effectively [29][30]. This summary encapsulates the key points discussed in the conference call, providing insights into the growth potential and strategic direction of the Scale Up network within the computing power industry.