量子计算

Search documents
美股面临“三连阴”压力 经济数据拖累金铜走弱|今夜看点
Sou Hu Cai Jing· 2025-09-29 23:41
智通财经9月25日讯(编辑 史正丞)连跌了两天的美股市场,今天晚上也很有可能以弱势姿态开盘。显然鲍威尔那句"股票估值相当高"仍然余音绕梁,市场 要等待新的催化因素才能翻篇。 (纳斯达克指数日线图,来源:TradingView) 截至发稿,纳斯达克100指数期货(2512合约)跌0.67%,标普500指数期货跌0.47%,道指期货跌0.25%。 发稿前刚刚公布的最新经济数据也是喜忧参半。一方面 美国第二季度实际GDP(年化季率)增长3.8%,明显高于预期值3.3%,高频初请失业人数也低于预 期;但 二季度核心PCE物价指数年化季率终值达到2.6%,同样高于预期。最新数据对股指期货影响不大,但 国际金价盘前明显跳了一波水,铜、原油等大 宗商品也受到影响。 (现货黄金分钟线图,来源:TradingView) 对于本周而言,投资者接下来要等待周五的月度PCE数据,另外距离美国政府停摆的关键节点也日益临近。 Polymarket平台的最新数据显示, 目前押注"美国政府在2025年关门"的概率已经升至76%,而10月1日前关门的概率为63%。 (以太坊日线图,来源:TradingView) 另外据衍生品交易所Deribi ...
腾讯研究院AI速递 20250930
腾讯研究院· 2025-09-29 16:01
Group 1: Generative AI Developments - DeepSeek-V3.2-Exp introduces Sparse Attention mechanism, significantly improving long text training and inference efficiency without compromising performance [1] - The model is open-sourced on HuggingFace and Modao platforms, with accompanying papers and code released [1] - Official API prices have been reduced by over 50% due to decreased service costs, with V3.1-Terminus interface available until October 15 for comparison [1] Group 2: RoboBrain-X0 Innovations - RoboBrain-X0 achieves zero-shot cross-ontology generalization, allowing deployment on various real robots with just pre-training [2] - The core innovation focuses on learning "what to do" rather than "how to move," standardizing complex actions into token sequences [2] - In real-world cross-ontology evaluations, the overall success rate reached 48.9%, nearly 2.5 times that of the baseline model π0, with a 100% success rate in basic grasping tasks [2] Group 3: 3D Generation Breakthroughs - The 3D-Omni model is the first to unify multiple conditional controls for 3D generation, supporting various control signals [3] - It employs a lightweight unified control encoder and progressive difficulty-aware training strategy for detailed 3D asset generation [3] - The model effectively addresses the "paper object" issue in single-view generation, accurately reconstructing geometric details and proportions [3] Group 4: Quantum Computing Advances - Caltech team sets a new record with a quantum bit array of 6100 qubits, achieving a coherence time of 13 seconds and a single-qubit control precision of 99.98% [6] - The team utilized optical tweezers to capture atoms and move qubits while maintaining superposition, highlighting the advantages of neutral atom systems over superconducting circuits and ion traps [6] - This achievement balances scale, precision, and coherence, reinforcing neutral atoms as a leading platform for quantum computing, though large-scale error correction demonstrations are still needed for practical applications [6] Group 5: AI Integration Predictions - Julian Schrittwieser from AlphaGo argues against the notion of AI stagnation, emphasizing significant advancements in AI capabilities over recent years [7] - METR research indicates exponential growth in AI abilities, with the latest models capable of autonomously completing tasks over two hours, and a trend of doubling capabilities every seven months [7] - Predictions suggest that by mid-2026, models may autonomously work for eight hours, achieving expert-level performance across multiple industries by the end of the year [7] Group 6: GPU Market Dynamics - The dominance of NVIDIA GPUs is expected to be challenged within 2-3 years as specialized chips for different workloads emerge, shifting the market from a 90% concentration to a more diversified ecosystem [8] - Inference costs have decreased by 100 times and may drop another 10 times, driven by advancements in MoE architecture, model quantization, and collaborative design between algorithms and hardware [8] - AI applications are anticipated to diversify into three categories: traditional chatbots, ultra-low latency scenarios, and large-scale batch processing, with hardware suppliers needing to optimize accordingly [8]
美股年内创下28次新高 标普迈向7000点? | 华尔街观察
Di Yi Cai Jing· 2025-09-29 15:37
Core Viewpoint - The U.S. stock market has shown resilience, with the S&P 500 index achieving a return of +2.3% in September, significantly above the historical average of -0.6%, driven by the AI wave and positive market sentiment [1][2]. Group 1: Market Performance and Predictions - The S&P 500 index is projected to reach levels above 7000 points, with Goldman Sachs raising its return forecasts for the next 3, 6, and 12 months to +2%, +5%, and +8%, corresponding to index levels of approximately 6800, 7000, and 7200 points [1][2]. - Morgan Stanley maintains a more conservative target of 6500 points but acknowledges the increasing likelihood of a bullish scenario reaching 7200 points [1][2]. Group 2: Earnings Growth and Market Dynamics - Earnings growth remains a key support for the bull market, with 54% of the S&P 500's year-to-date increase attributed to earnings growth, 38% to valuation expansion, and 8% to dividends [3]. - The actual yield on 10-year U.S. Treasuries has decreased from 2.23% to 1.75%, supporting stock market valuations [3]. Group 3: AI Industry as a Driving Force - The AI industry is identified as the core driver of the current bull market, with its extensive ecosystem expected to sustain upward momentum in the overall index [4][5]. - Recent strategic partnerships, such as NVIDIA's collaboration with OpenAI, have significantly boosted market sentiment and stock prices, with OpenAI's valuation soaring to $500 billion from $15 billion two years ago [5][6]. Group 4: Investment Sentiment and Strategies - Market participants express a cautious optimism regarding the AI sector, noting that while many AI stocks appear overvalued in the short term, the fundamental backdrop remains attractive due to relatively cheap capital and the Federal Reserve's easing stance [6][7]. - Experienced traders are advised to wait for potential market corrections before entering positions, contrasting with less experienced traders who may chase high prices [7].
美股年内创下28次新高
Di Yi Cai Jing Zi Xun· 2025-09-29 12:23
本文字数:2344,阅读时长大约4分钟 作者 |第一财经 韦薇 2025.09.29 三季度行将结束,今年美股已有28次创下历史新高。截至2025年9月25日,标普500指数9月(9月1日至9 月25日)的回报率为 +2.3% ,显著高于历史平均的-0.6%,并创下9月回报率新高。 市场人士纷纷感叹:"世界这么乱,股市怎么还涨?"尤其是9月向来是美股季节性最差的一个月,但在 人工智能(AI)浪潮驱动下,美股逆势冲高,多家华尔街投行给出了标普500指数7000点以上的目标 位。美股四季度何去何从?AI主题能支撑大盘多久?美联储连续降息的预期能否实现? 标普500指数迈向7000点? 目前,多家华尔街投行持更为积极的观点。例如,年内高盛已经几度调升目标价,目前将标普500未来3 个月、6个月、12个月的回报预测分别调升至+2%、+5%、+8%,对应指数水平约6800点、7000点、 7200点。 尽管摩根士丹利仍维持6500点的基准情景假设,但该机构首席美股策略师威尔逊(Michael Wilson)多 次表示,7200点的牛市情景越发有可能实现。 威尔逊在发给第一财经的邮件中提及,劳动力数据疲软的局面是已过去还 ...
【太平洋科技-每日观点&资讯】(2025-09-30)
远峰电子· 2025-09-29 11:56
行情速递 ①主板领涨,盈趣科技(+10.03%)/能科科技(+10.01%)/冠石科技(+10.00%)/领益智造 (+9.99%)/德明利(+8.00%)/ ②创业板领涨,汇金股份(+20.00%)/初灵信息(+20.00%)/凯旺科技(+18.30%)/ ③科创板领涨,品茗科技(+20.01%)/昀冢科技(+15.25%)/中科通达(+12.94%)/ ④活跃子行业,SW半导体设备(+3.41%)/SW印制电路板(+3.16%)/ 国内新闻 ①艾邦ARAI眼镜资讯,重庆康佳光电外延芯片团队在AR微显示领域取得重要 进展/团队依托单片集成技术/研制出主动驱动型蓝光Micro LED显示屏/实现 640×480分辨率、3.75μm像素间距和6773 PPI像素密度/ ②半导体产业纵横,兆芯官网公布了开胜KH-50000系列服务器处理器的更 多规格细节/该服务器处理器基于Chiplet架构/单颗处理器最多集成96个高性 能计算核心/提供高达384MB三级缓存/ ③半导体芯闻,北极雄芯QM935-G1成功测试点亮/该IVI Chiplet(车载信 息娱乐系统)专为智能座舱设计/包含高性能GPU核以及HIFI5、U ...
Is IBM Stock a Buy on Its Quantum Breakthrough?
Yahoo Finance· 2025-09-29 10:45
Key Points IBM and HSBC launched the world's first quantum-enabled algorithmic trading system in September, moving quantum computing from lab experiments to live financial markets. IBM has booked over $1 billion in cumulative quantum business since 2017, while maintaining profitable operations and a high dividend yield. Unlike pure plays trading at hundreds of times sales, IBM offers quantum exposure through a diversified business generating $13.5 billion in annual free cash flow. 10 stocks we like ...
创新药、低空飞行器、量子计算、脑机接口 院士报告为颠覆性技术“把脉问诊”
Ke Ji Ri Bao· 2025-09-29 09:48
Group 1: Innovation in Pharmaceuticals - China has made significant progress in innovative drug development, with the number of new drug approvals increasing from 4 in 2015 to 37 in 2024, and the proportion of first-in-class drugs rising from 4% to 38% [1] - Despite the advancements, challenges such as lack of original theories and technologies, intense competition, changing international environments, and payment system issues pose risks to sustainable industry growth [1][2] Group 2: Development of eVTOL Industry - The eVTOL industry in China is rapidly developing, currently leading globally, supported by the foundations laid by the aviation and new energy vehicle industries [2] - Key technological challenges include distributed propulsion aerodynamics, high-energy power sources, and safety control in complex environments, with a focus on smart technologies to enable autonomous flight [2] Group 3: Quantum Computing and Technology - The development of quantum technology is highlighted as a path for differentiation, with significant progress in solid-state quantum computing and scientific instrument research [3] - Emphasis is placed on the need for China to strengthen its independent research and development of scientific instruments to become a technological powerhouse [3] Group 4: Brain-Computer Interface Advancements - Brain-computer interfaces are categorized into brain control and brain modulation, with potential applications in aiding patients with movement disorders and treating mental health conditions [3] - Innovative non-invasive techniques are being explored to target specific brain areas for treatment, avoiding the side effects associated with traditional invasive surgeries [3] Group 5: Overall Technological Landscape - The reports from the four academicians reflect China's technological self-reliance and the competitive landscape of global technology [4] - Addressing the challenges is crucial for China to maintain a leading position in disruptive innovation and to secure control over technological advancements [4]
国盾量子(688027):量子计算驱动业绩高增,与中电信协同效应显著
Haitong Securities International· 2025-09-29 08:01
Investment Rating - The report maintains an "Outperform" rating for the company [4][8]. Core Insights - Quantum computing drives significant revenue growth, with a 74.54% year-on-year increase in H1 2025, and quantum computing revenue specifically rising by 283.92% [9]. - The synergy with China Telecom is becoming increasingly significant, transitioning quantum communication into an information security infrastructure, which is expected to enhance user convenience and accelerate industry growth [10]. - The commercialization of quantum computing is accelerating, highlighted by the successful delivery of the 504-qubit superconducting quantum computer "Tianyan-504" to China Telecom and a breakthrough in overseas markets with a 25-qubit delivery [11]. Financial Summary - The projected revenue for the company from 2025 to 2027 is estimated at 0.36 billion, 0.49 billion, and 0.64 billion RMB respectively, with earnings per share (EPS) expected to be 3.54, 4.80, and 6.22 RMB [4][8]. - In H1 2025, the company reported a net profit attributable to shareholders of -23.79 million RMB, with a recurring net profit of -46.75 million RMB [9]. - The company’s total revenue for 2023 is reported at 156 million RMB, with a projected increase to 640 million RMB by 2027, reflecting a compound annual growth rate (CAGR) of approximately 29.7% [3].
GPT-5攻克「量子NP难题」,首篇论文引爆学界,人类2周压缩至30分钟
3 6 Ke· 2025-09-29 03:19
Core Insights - GPT-5 has demonstrated the ability to solve complex scientific problems, including a "quantum version of the NP problem," in just 30 minutes, a task that would typically take humans 1-2 weeks to complete [1][7][19] - The AI's involvement in the research process has sparked discussions about the potential for AI to achieve significant breakthroughs in scientific fields, possibly even earning Nobel Prize-level recognition in the future [1][19] Group 1: Research Breakthrough - Scott Aaronson, a quantum computing expert, published a paper showing that GPT-5 assisted in solving a long-standing problem in quantum computing related to the QMA complexity class, which is akin to the NP problem [4][10] - The research revealed that existing methods had reached a theoretical limit of "double-exponential error," and GPT-5 provided a novel mathematical function that became crucial for the breakthrough [6][18] Group 2: AI's Role in Research - The interaction between Scott and GPT-5 involved several iterations, where the AI initially provided incorrect suggestions but ultimately refined its approach to yield a successful solution [21][22] - Scott noted that while GPT-5 cannot yet independently write entire research papers, it can significantly assist researchers in overcoming complex challenges, indicating a promising application for AI in scientific research [23]
IBM是“真量子龙头”,但市场对汇丰新闻反应过度
Hua Er Jie Jian Wen· 2025-09-29 01:37
一则关于量子计算商业化应用的新闻,让科技巨头IBM的股价坐上了"火箭"。 市场狂热:一日暴涨130亿背后的"估值泡沫" 量子计算"首秀":汇丰应用成果显著 报告首先引述了汇丰银行的新闻稿。据该新闻稿,汇丰银行利用IBM的Heron量子处理器,在预测欧洲公司债券市场的客户询价中标概率方面,与 传统计算资源相比,实现了高达34%的性能提升。 汇丰银行指出,这一结果展示了量子计算机在解决金融服务业动态问题时可能提供的价值。报告认为,这类早期的商业应用对于证明量子计算的 价值主张"极其重要"。 IBM:无可争议的量子霸主 报告明确表示,IBM是量子市场的"明确领导者"。研究表明,IBM拥有当今最大、最广泛的先进量子计算机生态系统。 在汇丰银行宣布成功使用IBM的量子处理器优化债券交易后,IBM的市值在9月25日当天增加了近130亿美元,股价当天涨超5%。 然而,据追风交易台消息,摩根士丹利分析师Erik W Woodring团队发布的报告认为,这一市场反应虽然凸显了IBM在量子领域的领导地位,但也 可能过度消化了这项仍处于早期阶段的技术的未来潜力。 该行通过模型反向推算指出,市场当天的狂热反应,相当于预期IBM在20 ...