Core Viewpoint - The article highlights the significant advancements in China's supercomputing capabilities, particularly through the contributions of Meng Xiangfei, emphasizing the journey from initial struggles to becoming a global leader in supercomputing technology [2][4]. Group 1: Supercomputing Development - The new generation "Tianhe" supercomputer can perform calculations equivalent to 7 billion people using an abacus for 10,000 years, showcasing its immense computational power [2]. - In 2010, "Tianhe-1" topped the world supercomputer rankings, marking a pivotal moment for China's supercomputing industry [4][6]. - The development of supercomputing in China involved overcoming significant technical challenges and a commitment to national pride and innovation [2][4]. Group 2: Personal Journey of Meng Xiangfei - Meng Xiangfei's journey from a theoretical physicist to the chief scientist of the National Supercomputing Center in Tianjin reflects a deep sense of national responsibility and commitment to innovation [4][5]. - His decision to return to China in 2009 was influenced by a strong sense of patriotism, coinciding with the critical period of "Tianhe-1" development [4][5]. Group 3: Overcoming Challenges - Despite achieving the top ranking, "Tianhe-1" faced skepticism from international peers, who viewed it as merely a "gaming machine" [6]. - Meng Xiangfei and his team worked tirelessly to demonstrate the practical applications of their technology, significantly reducing the time required for complex tasks in oil exploration from months to just 16 hours [6]. Group 4: Acknowledgment of Collective Efforts - The rise of Chinese supercomputing is attributed to the collective efforts of many individuals, not just one person, highlighting the importance of teamwork and perseverance [8][9]. - Meng Xiangfei honors the contributions of countless engineers and researchers who have dedicated their lives to advancing supercomputing in China [9]. Group 5: Future Vision - Meng Xiangfei envisions a future where China's supercomputing capabilities will expand into new areas, including quantum computing, and will have a global presence [10][11]. - He emphasizes the importance of continuous learning and innovation for the younger generation to thrive in the age of artificial intelligence [11].

11月17日，评测超级计算机计算能力的权威排行榜——第66届国际超算TOP500榜单在SC25大会上正式 揭晓。本次榜单最大亮点在于，德国于利希超算中心的"JUPITER Booster"完成全面部署后，性能成功 突破E级（百亿亿次/秒）大关，不仅成为欧洲首台E级超算，更是TOP500榜单中的第四台E级超算系 统。整体来看，本届榜单TOP10阵营保持稳定格局，美国能源部旗下的El Capitan、Frontier、Aurora三 台E级超算继续包揽前三，无新系统跻身前十。 | Rmax | Power | Site | Rank | Computer | Contury | Cores | PFlop/s) | (KW) | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | El Capitan | DOE/NNSA/LLNL | HPE Cray EX255a, AMD 4th Gen EPYC 24C 1.8GHz, | USA | 1 | 11,039,616 | 1,809.00 | 29,685 | | | AMD Instinct ...

Core Insights - Eviden has launched the BullSequana XH3500, a supercomputer designed to support the convergence of High-Performance Computing (HPC) and Artificial Intelligence (AI) workloads, aiming for exascale performance [1][2] - The BullSequana XH3500 enhances performance, energy efficiency, and scalability, integrating a comprehensive ecosystem of software and services to accelerate scientific discovery and industrial innovation [3][4] Performance and Efficiency - The BullSequana XH3500 provides over 80% more electrical power per square meter and 30% greater cooling capacity per kilowatt compared to its predecessor, enabling industry-leading performance density and energy efficiency [4] - The system features 100% fan-less 5th generation Direct Liquid Cooling (DLC) with warm water support, significantly reducing total cost of ownership (TCO) and supporting greener operations [6] Flexibility and Modularity - The open, modular design of the BullSequana XH3500 allows customers to mix and match different computing, acceleration, and networking technologies, ensuring full technology freedom and eliminating vendor lock-in [5][7] - This modular architecture supports diverse CPU, GPU, and AI accelerators, enabling tailored configurations for various applications, including large-scale AI training [7] Company Overview - Eviden, part of the Atos Group, generates approximately €1 billion in revenue and operates in 36 countries, focusing on advanced computing, cybersecurity, mission-critical systems, and vision AI [8] - Atos Group, with around 67,000 employees and annual revenue of approximately €10 billion, is a leader in digital transformation, committed to a secure and decarbonized future [9]

Core Insights - The inauguration of the JUPITER Booster, an AI-dedicated partition, marks a significant advancement in high-performance computing (HPC) and AI capabilities in Europe [1][2][4] - JUPITER is recognized as Europe's most powerful HPC and AI system, with the ability to perform over one quintillion calculations per second, equivalent to the power of 10 million modern desktop computers [2][3] - The JUPITER system integrates advanced technologies, including 24,000 NVIDIA GH200 Grace Hopper Superchips and Direct Liquid Cooling technology, achieving exceptional energy efficiency and performance [4] Company Overview - Eviden, a brand of Atos Group, specializes in advanced computing, cybersecurity products, mission-critical systems, and vision AI, generating approximately €1 billion in revenue and operating in 36 countries [7] - Atos Group, with around 70,000 employees and annual revenue of approximately €10 billion, is a leader in digital transformation and is committed to a secure and decarbonized future [8] Technological Impact - JUPITER's capabilities will enhance research in various fields, including climate science, sustainable energy, and AI, enabling unprecedented innovation across Europe [5][13] - The supercomputer's power allows for detailed simulations in climate modeling and brain research, significantly improving understanding of extreme weather events and neurodegenerative diseases [13]

Group 1 - Tesla's vice president of hardware design engineering, Pete Bannon, is leaving the company after joining in 2016 from Apple, and he was leading the development of Tesla's Dojo supercomputer [1] - Musk has been positioning Tesla as an artificial intelligence and robotics powerhouse, with Dojo being a key component designed to process and train AI models using data from Tesla vehicles [2] - The focus on Dojo and another computing cluster, Cortex, aims to enhance Tesla's advanced driver assistance systems and fulfill Musk's vision of transforming existing Teslas into robotaxis [3] Group 2 - Tesla has entered a $16.5 billion deal with Samsung to produce A16 chips domestically, and is currently testing a Robotaxi service in Austin, Texas, which requires a human safety supervisor [4] - During an earnings call, Musk addressed concerns about competition for AI talent between Tesla and his AI company, xAI, stating that both companies are focused on different aspects of AI development [5][6] - Tesla has seen a significant departure of top talent this year, including key figures in robotics and software engineering, indicating potential challenges in retaining skilled personnel [7]