计算机

Search documents
易普森科技取得基于计算机视觉技术的机器人控制方法及系统专利
Sou Hu Cai Jing· 2025-05-16 03:50
易普森生物科技(深圳)有限公司,成立于2017年,位于深圳市,是一家以从事批发业为主的企业。企 业注册资本1000万人民币。通过天眼查大数据分析,易普森生物科技(深圳)有限公司共对外投资了1 家企业,参与招投标项目73次,专利信息47条,此外企业还拥有行政许可42个。 来源:金融界 天眼查资料显示,深圳易普森科技股份有限公司,成立于2011年,位于深圳市,是一家以从事专业技术 服务业为主的企业。企业注册资本6148.9058万人民币。通过天眼查大数据分析,深圳易普森科技股份 有限公司共对外投资了7家企业,参与招投标项目127次,财产线索方面有商标信息16条,专利信息70 条,此外企业还拥有行政许可17个。 易普森智慧健康科技(深圳)有限公司,成立于2017年,位于深圳市,是一家以从事软件和信息技术服 务业为主的企业。企业注册资本3000万人民币。通过天眼查大数据分析,易普森智慧健康科技(深圳) 有限公司共对外投资了1家企业,参与招投标项目126次,财产线索方面有商标信息4条,专利信息134 条,此外企业还拥有行政许可20个。 金融界2025年5月16日消息,国家知识产权局信息显示,深圳易普森科技股份有限公司、 ...
神州泰岳:2024年报及2025年一季报点评Q4符合、Q1有所承压,期待新产品提供流水增量-20250515
Huachuang Securities· 2025-05-15 00:35
公司研究 证 券 研 究 报 告 神州泰岳(300002)2024 年报及 2025 年一季报点评 推荐(维持) Q4 符合、Q1 有所承压,期待新产品提供流 水增量 事项: ❖ 公司发布 24Q4&25Q1 财报:Q4 符合,Q1 有所承压。其中 4Q24 业绩落于前 期预告区间中枢附近,单季度实现收入 19.4 亿元,YOY+1.5%,QOQ+33%; 实现归母净利润 3.3 亿元,YOY+3%,QOQ-28%;实现扣非归母净利润 3.1 亿 元,YOY+7%,QOQ+46%。1Q25 业绩有所承压,单季度实现收入 13.2 亿元, YOY-11%,QOQ-32%;实现归母净利润 2.4 亿元,YOY-20%,QOQ-29%;实 现扣非归母净利润 2.3 亿元,YOY-21%,QOQ-28%。24 年年末现金分红 1.96 亿元,分红率 14%,YOY 基本稳定。 评论: [ReportFinancialIndex] 主要财务指标 | | 2024A | 2025E | 2026E | 2027E | | --- | --- | --- | --- | --- | | 营业总收入(百万) | 6,452 ...
神州泰岳(300002):2024年报及2025年一季报点评:Q4符合、Q1有所承压,期待新产品提供流水增量
Huachuang Securities· 2025-05-14 13:13
公司研究 证 券 研 究 报 告 神州泰岳(300002)2024 年报及 2025 年一季报点评 推荐(维持) Q4 符合、Q1 有所承压,期待新产品提供流 水增量 事项: ❖ 公司发布 24Q4&25Q1 财报:Q4 符合,Q1 有所承压。其中 4Q24 业绩落于前 期预告区间中枢附近,单季度实现收入 19.4 亿元,YOY+1.5%,QOQ+33%; 实现归母净利润 3.3 亿元,YOY+3%,QOQ-28%;实现扣非归母净利润 3.1 亿 元,YOY+7%,QOQ+46%。1Q25 业绩有所承压,单季度实现收入 13.2 亿元, YOY-11%,QOQ-32%;实现归母净利润 2.4 亿元,YOY-20%,QOQ-29%;实 现扣非归母净利润 2.3 亿元,YOY-21%,QOQ-28%。24 年年末现金分红 1.96 亿元,分红率 14%,YOY 基本稳定。 评论: [ReportFinancialIndex] 主要财务指标 | | 2024A | 2025E | 2026E | 2027E | | --- | --- | --- | --- | --- | | 营业总收入(百万) | 6,452 ...
新技术精确模拟银河系星际湍流
Ke Ji Ri Bao· 2025-05-13 23:26
Core Insights - A collaborative team from the University of Toronto, Princeton University, and the Australian National University has developed an innovative computer simulation technology that explores the interstellar medium (ISM) with unprecedented precision and scale [1][2] - The new model provides insights into astrophysical phenomena such as ISM, galactic magnetic fields, star formation, and cosmic ray propagation, and has been published in the latest issue of Nature Astronomy [1] Group 1: Model Capabilities - The new model achieves significant breakthroughs in size and detail, capable of simulating a spatial volume approximately 30 light-years across, with the smallest version reduced to about 1/5000 of that size [2] - It offers higher resolution and scalability, allowing for the simulation of dynamic changes in ISM density, which previous models could not account for [2] Group 2: Scientific Implications - The model utilizes observational data from the solar-Earth system to validate simulation results, indicating a better understanding of space weather and its effects on Earth [2] - Turbulence is a common phenomenon in various natural and engineering contexts, and the model enhances understanding of how magnetic fields influence gas flow and star formation, as well as the impact of cosmic weather events like solar storms on Earth [2]
长春光机所光电突触器件研究取得新进展
Huan Qiu Wang Zi Xun· 2025-05-10 09:18
来源:中国新闻网 中新网长春5月10日电 (记者 郭佳)记者10日从中国科学院长春光学精密机械与物理研究所(下称"长春光 机所")获悉,该所科研人员开发出一种紫外光电突触器件和一种光电突触晶体管,这两项成果分别为先 进人工视觉系统和神经形态计算视觉的发展提供了新的技术路径。 紫外光电突触器件由长春光机所特种发光科学与技术全国重点实验室黎大兵、孙晓娟研究员领导的研究 团队成功开发。他们通过巧妙的器件设计,利用AlScN的铁电极化特性和GaN的优异光电性能,基于异 质结处空穴的俘获与解俘获机制成功构建了一种新型的紫外光电突触器件。 该器件不仅具备出色的非易失性存储特性,还能够模拟生物视觉系统中的突触功能,进行多阻态调节, 如长时程增强(LTP)、对脉冲增强(PPF)以及学习-遗忘-再学习过程。 光电突触晶体管由该所李绍娟、黎大兵领导的研究团队完成,通过气体吸附辅助的持久光电导策略,实 现从紫外到近红外的宽谱高光电转换效率和长时数据保留能力。 该器件在紫外到近红外波段展现出优异的光电探测性能;通过气体吸附辅助的持久光电导策略,器件在 375nm-1310nm宽谱范围内展现出长时数据保留能力,双脉冲易化(PPF)指数 ...
云计算沪港深ETF(517390)大涨超4%,位居ETF涨幅榜前三
Xin Lang Cai Jing· 2025-05-06 06:47
截至2025年5月6日 14:25,中证沪港深云计算产业指数(931470)强势上涨2.74%,成分股天源迪科(300047)上涨20.00%,华胜天成(600410)上涨10.01%,润和 软件(300339)上涨9.01%,数据港(603881),中国软件国际(00354)等个股跟涨。云计算沪港深ETF(517390)上涨4.16%,位居ETF涨幅榜前三。最新价报1.13 元。拉长时间看,截至2025年4月30日,云计算沪港深ETF近2周累计上涨3.15%。流动性方面,云计算沪港深ETF盘中换手4.66%,成交1382.15万元。拉长时 间看,截至4月30日,云计算沪港深ETF近1月日均成交994.03万元。 截至4月30日,云计算沪港深ETF近3年净值上涨56.99%,指数股票型基金排名12/1742,居于前0.69%。从收益能力看,截至2025年4月30日,云计算沪港深 ETF自成立以来,最高单月回报为33.39%,最长连涨月数为6个月,最长连涨涨幅为66.14%,上涨月份平均收益率为9.82%,年盈利百分比为66.67%,历史 持有3年盈利概率为80.85%。截至2025年4月30日,云计算沪港深E ...
数读科创板药企年报:过半净利增长,平均研发费用超3.5亿
Bei Jing Shang Bao· 2025-05-05 14:17
科创板已成为生物医药产业聚集地,截至目前,科创板药企2024年年报出齐。Wind数据显示,目前科创板共有115家生物医药企业,2024年,有超过半数药 企净利同比出现增长。其中微电生理(688351)净利增幅最猛,同比增逾8倍。多家"U"标企业首次实现盈利,成功摘"U"。研发能力是科创板药企竞争力的 重要体现,2024年,115家科创板药企平均研发费用为3.53亿元,其中百济神州研发费用"一骑绝尘",为141.4亿元。不过,2024年,有部分医药股研发费用 出现大幅下降,其中科兴制药研发费用同比"腰斩"。此外,荣昌生物等医药股研发人员数量出现大幅下降。 | 序号 | 证券代码 | > 证券简称 | レ ا | レ 归属母公司股东的净利润 | 归属母公司股东的 | | --- | --- | --- | --- | --- | --- | | | | | [报告期] 去年年报 [报表类型] 合并报表 | | [报告期] 去年年捉 [单位] % | | | | | [单位] 亿元↓ | | | | 1 | 688506 SH | 百利天恒 | 3 | 37.0750 | | | 2 | 688578.SH | 艾 ...
神州泰岳(300002):存量游戏稳健运营 静待新游周期开启
Xin Lang Cai Jing· 2025-04-29 02:48
Core Viewpoint - The company reported a revenue of 6.452 billion yuan in 2024, representing a year-over-year increase of 8.22%, with a net profit of 1.428 billion yuan, up 60.92% year-over-year, indicating strong financial performance despite a decline in Q1 2025 [1] Group 1: Game Business - The gaming segment achieved a revenue of 4.662 billion yuan in 2024, with a year-over-year growth of 3.66% and a gross margin of 72.70%, reflecting a 1.56 percentage point increase [2] - The net profit from gaming was 1.552 billion yuan, up 40.47% year-over-year, with a net profit margin of 33.28%, an increase of 8.72 percentage points [2] - The company expects a natural decline in revenue from existing products in Q1 2025, leading to an overall revenue decrease of 11.07% year-over-year [2][3] Group 2: Future Outlook - The company plans to maintain long-term operations for existing games, with expectations for stable revenue from key titles like "Age of Origins" and "War and Order" [3] - New games "Stellar Sanctuary" and "Next Agers" are set to launch in overseas markets by the end of 2024, with plans for large-scale promotion in mid to late 2025 [3] - The company has additional RPG and SLG games in the pipeline, which are anticipated to contribute to revenue growth [3] Group 3: Computer Business - The computer and other businesses generated revenue of 1.790 billion yuan in 2024, marking a year-over-year increase of 22.21% [4] - AI/ICT operations accounted for 1.478 billion yuan, growing 20.53% year-over-year, while IoT/communications revenue reached 133 million yuan, up 62.30% [4] - The company is focusing on integrating AI across various business lines, enhancing profitability and growth potential [4] Group 4: Profit Forecast and Investment Rating - The company has adjusted its earnings per share (EPS) forecasts for 2025-2027 to 0.59, 0.64, and 0.69 yuan, respectively, reflecting a downward revision from previous estimates [5] - The current price-to-earnings (PE) ratios are projected at 19, 18, and 17 times for 2025-2027 [5] - The company maintains a "buy" rating, supported by strong computer business growth and a robust pipeline of new games [5]
计算机指数偏离修复模型效果点评
Tai Ping Yang Zheng Quan· 2025-04-28 15:33
- Model Name: Deviation Correction Model for Computer Index[4] - Model Construction Idea: The model assumes that the price trend of the target relative to the reference target has a repeated deviation-regression cycle, and the degree of deviation has a limit. By statistically analyzing the existing data, a reasonable threshold can be found. When the price is below this value, it indicates that the current price deviation is close to the limit, and the space for further decline is limited, so it is suitable to buy and hold until the price returns[6] - Specific Construction Process: 1. Calculate the closing price of the SW First-level Computer Index relative to the CSI 300 Index within the calculation period, denoted as cl[5] 2. Calculate the drawdown sequence W within the period[5] 3. Calculate the maximum value during each single drawdown period in W[6] 4. Arrange the maximum drawdown values in descending order to get sequence S, and sum S to get T[6] 5. Perform cumulative addition on S, stop when the cumulative sum reaches or exceeds 80% of T, and remove the values in S that did not participate in the cumulative addition, resulting in sequence S1[6] 6. Use an iterative method to further screen the elements in S1 by repeating the above two steps until the iterative result no longer changes, resulting in the effective drawdown set C[6] 7. Take 80% of the maximum value in C as the threshold. When W is greater than this threshold, the signal is 1 (buy); when W is 0, the signal is also zero (close position); in other cases, the signal is set to the previous value[6] - Model Evaluation: The model is not applicable in the tracking period as it shows significant volatility and long-term drawdown, with the final drawdown exceeding the previous ones, indicating that the current price trend is outside the statistical sample[7] Model Backtest Results - Total Strategy Return: 78.50%[6] - Buy and Hold Return: 152.35%[6] - Total Excess Return: -73.86%[6] - Maximum Drawdown: 54.68%[6] - Longest Drawdown Duration: 2261 trading days[6]