社区观点
XAUUSD M45:在OB 4.331–4.339处等待SELL,在支撑4.306–4.312处BUY1) 市场背景 (M45) – SMC & 价格结构
价格急升后大幅下跌是一个流动性事件(清扫流动性),形成明显的看跌位移。
在下跌之后,价格正在以回调/重测的方式回升 → 优先策略是“在供应区SELL回调”,或者“在需求区BUY反应”如果有下探。
2) 关键水平
流动性卖出(流动性高点):4.367.982
OB看跌(卖出区):4.331.123 – 4.338.610
流动性支撑(需求):4.312.463 – 4.306.358
当前参考价格:~4.326(位于区间中间,不是最佳入场点)
3) 交易计划
情景A – 在OB看跌处SELL(优先)
如果价格回升至4.331 – 4.339区域并出现拒绝信号:
pinbar / 下影线
吞没下降
或M15–M45上的ChoCH下降
此时优先SELL短期下降结构中的回调。
参考目标:
TP1: 4.312
TP2: 4.306
无效化:
M45蜡烛坚实收于4.339之上并保持 → 停止SELL想法。
情景B – 在流动性支撑处BUY(逆转情景)
如果价格被拉至4.312 – 4.306区域并有清扫+回收迹象(下探后重新收回区域)。
确认后,等待BUY回调。
参考目标:
TP1: 4.331
TP2: 4.339
注意:如果价格再次触及OB看跌而未能突破,优先平仓并观察反应。
情景C – 上探流动性卖出4.368后再逆转
如果价格突破OB看跌并继续流动性上升至4.368。
4.368区域是寻找卖出反应的合适位置(仅在有确认信号时SELL)。
4) 18/12影响黄金的新闻
12月18日有美国CPI(11月):此类新闻对黄金影响最大,因为直接影响利率预期、收益率和美元。
同日通常有初请失业金人数和活动指标(如费城联储),容易为XAUUSD制造短期波动。
CPI发布后,市场通常在选择方向前会有双向流动性清扫行为 → 限制FOMO,优先在正确区域交易。
5) 快速结论
短期偏向:优先在OB 4.331–4.339处SELL回调,直到价格强势回收上方。
防御情景:如果有清扫+确认,在4.312–4.306处BUY反应。
避免在区间中间入场;等待“正确区域 – 正确信号”。
XAU/USD – 流动性抓取已完成,等待回调继续市场背景
黄金刚刚完成了一次强劲的冲动性上涨,扫除了前高之上的买方流动性。在这一扩张之后,价格现在进入了短期整合/回调阶段,这是流动性抓取后的预期行为,而不是完全的趋势逆转。
从基本面来看,对美联储更宽松立场的预期继续限制黄金的下行压力,尽管日内有修正,但整体偏向仍然得到支持。
技术结构(短期)
较大的结构仍然看涨
最近的走势显示在高点处的流动性抓取,随后出现拒绝
当前的价格行为表明正在重新平衡进入需求区
目前尚未确认看跌市场结构的转变
关键技术区域
上方阻力/卖出反应:4,360 – 4,365
日内阻力:4,333 – 4,336
买入区(回调区域):4,300 – 4,305
更深的需求/流动性支持:4,275 – 4,280
交易计划 – MMF 风格
主要情景 – 买入回调
预计价格回撤至4,300 – 4,305或更深至4,280
寻找看涨反应/稳定以重新加入趋势
上行目标:
TP1:4,323
TP2:4,335
TP3:4,360+(流动性)
替代情景
如果价格保持在4,323以上并直接突破更高,等待清晰的重测后再考虑继续设置。
避免在高价位追逐价格。
无效
持续突破并在H1收盘低于4,275将削弱看涨延续情景,并将重点转向更深的整合。
总结
黄金仍处于看涨环境中,目前的弱势被视为流动性抓取后的技术性回调。
偏向仍然是逢低买入,优先考虑在关键需求区的耐心和确认。
12.19比特币行情本文撰写时间13:46
昨日比特币晚间先给到拉升,拉升高点在89468位置,随后半夜12点换线后快速下跌,下跌低点在84401附近,直线下跌5000点,以太坊拉升高点是2997附近,随后直线下跌至2772位置,反弹基本到达我们昨日布局目标,今日白盘关注高点破位;
昨日我们多单布局点位是86500附近,目标是89,拉升给到我们离场机会,利润3500点左右;拉升过后并未走出突破,随后4小时走出直线下跌,收大阴线,白盘下跌也未延续,走出小幅反弹,近期的走势都比较快,市场消息面可以看到多头被无限收割,这样的走势贪多就容易被收割;日线图上看,下方85500附近支撑大级别还未跌破,支撑还算有效,短线拉升需关注前面两天高点破位情况,昨日高点89460位置,能有效突破,后市将开始大级别反弹,短线上点关注上方88400这一点位到达情况,到达后4小时收上影线即可短线做空布局,止损推荐89500,目标先看85500附近,跌破84401持有到8万,如4小时并未收针就不去做空,关注高点98468位置突破,突破可追涨至91000附近,自行把握进场;
以太坊昨日多单布局点位2810附近,目标2950附近,反弹最高点2997位置,给到我们离场机会,利润140点左右;走势上跟大饼大同小异,拉升并未走出突破,下跌再度往2750去靠后白盘出拉升,以太日线5连阴后终于收阳,白盘交易上先关注2950附近,该位置到达收上影线可空,止损3000即可,目标看2800-2750,反之并未受上影线就不去布局,突破高点2997位置可追涨至3100-3150,自行把握进场机会;
短线交易,控制风险,盈亏自理;
原油 CRUDE 趋势与基本交易策略周线复合:混合,在技术因素同步之前,市场不太可能形成持续、稳定的趋势。
日线复合:多头,但仍需要日线⼀个超波多头反转的出现才能确认日线级别的低点已经形成。(牛)
关键因素:自10月⾼点 以来的下跌仍然属于修正结构,但原油已经运⾏⾄⼀个修正应当完成的极限区域。如果日线收盘价跌破 10 月的日线收盘低点,则将构成新的低点,并警告这轮下跌可能并非修正,⽽是趋势性的下跌。
交易策略建议:日线收盘价站上 12/10 的收盘价,是低点完成的初步信号;⽽明确的确认信号是收盘价突破12/5 的摆动⾼点收盘价。鉴于此前已经出现多次假反转与假突破,我会等待收盘价确认站上 12/5 的收盘价,再次建立多头仓位,以把握⼀轮可能持续数月的上涨⾏情。
⚠ 短期交易者应以TradingBox信号系统为准!
12.19 比特币走势分析:比特币仍处于C5浪的下跌中,即将进行C5-2浪的反弹,反弹继续做空(比特币合约交易)军长12.19 比特币价格今日行情:比特币目前下跌到了85,000附近,昨晚再次出现了典型的“画门”行情,快速拉升随后又快速回落,和前一天几乎一模一样。这种蹦极式走势,对交易并不友好,更多是在消耗情绪和耐心。
从细节上看,80,537这一线的反弹,是c四浪的结束点,之后出现的全部下跌,都属于c五浪内部结构。目前c五浪并不是简单的单边下跌,而是演化成了一个喇叭形的扩散结构,本质上仍然是一个下跌推动浪。接下来有条件出现一个二浪反抽,但这个反抽完成之后,依然是继续向下运行,直到整个c五浪彻底走完为止。
综上分析,比特币空单止盈,昨天比特有惊无险收益现货幅度2%或3.4%,ZEC收益现货幅度5%-6%;C5-2要不要做,先观察到晚上再决定。
一天一家A股公司高管:中科寒武纪科技股份有限公司董事长陈天石当资本成为诗行:一位企业掌舵者的价值投资叙事
一、基本信息
陈天石,男,1985年6月出生于江西南昌,汉族,中国国籍,无境外永久居留权。他是中科寒武纪科技股份有限公司(以下简称“寒武纪”)创始人、董事长兼CEO,同时也是中国科学院计算技术研究所研究员、九三学社中央委员会促进技术创新工作委员会副主任、第十四届全国政协委员。
二、教育背景
陈天石的教育经历贯穿“天才少年”与“学术深耕”的主线:
1998-2001年:就读于南昌市第二中学(初中);
2001-2005年:考入中国科学技术大学少年班,攻读数学与应用数学专业,获理学学士学位;
2005-2010年:保送至中国科学技术大学计算机学院,师从陈国良院士、姚新教授,攻读计算机软件与理论专业,获工学博士学位。
其教育经历体现了“早慧”与“专注”的特点:16岁进入中科大少年班,25岁完成博士学业,且在本科至博士阶段始终保持对计算机体系结构与人工智能的兴趣,为后续科研与创业奠定了基础。
三、职业经历
陈天石的职业路径分为“学术科研”与“创业实践”两大阶段,两者相互支撑,共同推动其在AI芯片领域的突破:
1. 学术科研阶段(2010-2016年)
博士毕业后,陈天石进入中国科学院计算技术研究所工作,历任助理研究员(2010-2013年)、副研究员(2013-2016年)、研究员(2016年至今)。期间,他专注于计算机体系结构与计算智能的交叉研究,重点探索AI芯片的基础理论与设计方法。
其学术成果丰硕,发表论文40余篇(涵盖IEEE/ACM Transactions、ASPLOS、ISCA、MICRO等国际顶级期刊与会议),并获得多项荣誉:
2014年,论文《DianNao:一种小尺度的高吞吐率机器学习加速器》获ASPLOS最佳论文奖,同年《DaDianNao》获MICRO最佳论文奖;
2011年,获中国计算机学会优秀博士论文奖、中国科学院优秀博士论文奖;
2012年,获全国百篇优秀博士论文提名奖。
这些成果奠定了他在AI芯片领域的学术地位,也为后续创业提供了技术积累。
2. 创业实践阶段(2016年至今)
2016年,陈天石带领团队核心成员创立寒武纪(原名“中科寒武纪科技股份有限公司”),并担任董事长兼CEO。公司名称“寒武纪”源自地质学上的“生命大爆发”时代,寓意“人工智能芯片将引发智能时代的物种大爆发”。
创业初期(2016-2019年):
2016年,推出世界首款终端人工智能专用处理器“寒武纪1A”,并集成至华为Mate10手机,实现AI芯片的商业化落地;
2017年,发布3款新品(能耗比超越“寒武纪1A”),覆盖图像识别、安防监控、智能驾驶等应用领域;
2018年,推出面向视觉领域的“寒武纪1H8”“寒武纪1H16”、面向智能驾驶的“寒武纪1M”,以及人工智能系统软件“Cambricon NeuWare”;
2019年,发布面向边缘人工智能计算的“思元220”芯片,形成“终端-边缘-云端”完整智能芯片产品线。
上市与扩张(2020年至今):
2020年7月20日,寒武纪在科创板上市(股票代码:688256),成为“国内AI芯片第一股”,首日开盘价250元/股,较发行价上涨288%,市值一度突破千亿;
2023年,推出“思元590”芯片(7nm工艺),推理能效比比肩国际巨头,支持国内主流大模型,迎来云端产品线爆发;
2024年,实现单季盈利,营收同比增长43倍;
2025年,寒武纪市值一度突破6000亿元,成为A股“科技龙头”之一。
社会任职:
2023年,当选第十四届全国政协委员、九三学社中央委员会促进技术创新工作委员会副主任;
2025年,获评“第六届全国非公有制经济人士优秀中国特色社会主义事业建设者”。
四、主要成就与荣誉
陈天石的成就涵盖学术研究与商业实践两大领域:
1. 学术成就
提出“算法-架构-芯片”协同设计理念,推动AI芯片从“通用”向“专用”转型;
发表多篇国际顶级论文(如ASPLOS、MICRO最佳论文),引领AI芯片领域的学术发展;
获得国家自然科学奖二等奖、中国科学院杰出科技成就奖等多项学术荣誉。
2. 商业成就
创立寒武纪,成为全球智能芯片领域的先行者,公司估值一度突破6000亿元;
推出多款具有自主知识产权的AI芯片(如“寒武纪1A”“思元590”),实现AI芯片的商业化落地,支持华为、小米等主流厂商的产品;
带领寒武纪成为“国产替代”的核心企业之一,在云端、边缘、终端等领域形成完整产品线。
五、个人特点与理念
陈天石的性格与理念贯穿“专注”“务实”“创新”三大关键词:
专注:深耕AI芯片领域十余年,拒绝短期盈利诱惑,坚持“死磕技术”;
务实:强调“芯片是硅疙瘩,要支持下游应用”,将学术成果转化为实际产品;
创新:提出“寒武纪大爆发”的理念,认为AI芯片将推动智能时代的到来,坚持“做难而正确的事”。
六、总结
陈天石从“天才少年”到“AI芯片创业者”的路径,体现了“学术深耕”与“商业实践”的结合。他创立的寒武纪,不仅推动了中国AI芯片产业的发展,也为全球智能时代提供了“中国方案”。其个人履历中的“教育背景”“学术成果”“商业成就”,共同构成了“科技创业者”的典型样本,激励着更多年轻人投身于硬科技领域。
Chen Tianshi
I. Basic Information
Chen Tianshi, male, born in June 1985 in Nanchang, Jiangxi Province, Han ethnicity, Chinese citizen with no permanent residence abroad. He is the founder, Chairman, and CEO of Cambricon Technologies Corporation Limited (hereinafter referred to as "Cambricon"), a researcher at the Institute of Computing Technology, Chinese Academy of Sciences (ICT), Vice Chairman of the Promotion Committee for Technological Innovation of the Central Committee of Jiusan Society, and a member of the 14th National Committee of the Chinese People's Political Consultative Conference (CPPCC).
II. Educational Background
Chen Tianshi's educational journey is characterized by "prodigy" and "academic dedication":
1998-2001: Studied at Nanchang No. 2 High School (junior high school).
2001-2005: Admitted to the University of Science and Technology of China (USTC) Young Talent Class, majoring in Mathematics and Applied Mathematics, and received a Bachelor of Science degree.
2005-2010: Recommended for postgraduate study at USTC's School of Computer Science and Technology, under the supervision of Academician Chen Guoliang and Professor Yao Xin, majoring in Computer Software and Theory, and received a Doctor of Engineering degree.
His academic path reflects "early maturity" and "persistence": entering USTC Young Talent Class at 16, completing his doctorate at 25, and maintaining a consistent interest in computer architecture and artificial intelligence from undergraduate to postgraduate studies, laying a solid foundation for his subsequent academic research and entrepreneurship.
III. Professional Experience
Chen Tianshi's professional career is divided into two interrelated stages: "academic research" and "entrepreneurial practice," both of which have promoted breakthroughs in the field of AI chips:
1. Academic Research Stage (2010-2016)
After graduating with a doctorate, Chen Tianshi joined ICT, serving successively as Assistant Researcher (2010-2013), Associate Researcher (2013-2016), and Researcher (2016-present). During this period, he focused on interdisciplinary research in computer architecture and computational intelligence, exploring the basic theories and design methods of AI chips.
His academic achievements are remarkable. He published over 40 papers in top international journals and conferences (including IEEE/ACM Transactions, ASPLOS, ISCA, MICRO, etc.) and received multiple honors:
2014: His papers DianNao: A Small-Footprint High-Throughput Accelerator for Ubiquitous Machine-Learningwon the ASPLOS Best Paper Award, and DaDianNaowon the MICRO Best Paper Award.
2011: Received the Outstanding Doctoral Dissertation Award of the China Computer Federation (CCF) and the Outstanding Doctoral Dissertation Award of ICT.
2012: Nominated for the National Outstanding Doctoral Dissertation Award.
These achievements established his academic status in the field of AI chips and provided technical support for his subsequent entrepreneurship.
2. Entrepreneurial Practice Stage (2016-Present)
In 2016, Chen Tianshi led the core team to establish Cambricon (formerly known as "Cambricon Technologies (Beijing) Co., Ltd.") and served as Chairman and CEO. The company name "Cambricon" is derived from the "Cambrian Explosion" in geology, implying that "AI chips will trigger a great explosion of intelligent species in the era of artificial intelligence."
Early Entrepreneurship (2016-2019):
2016: Launched the world's first terminal AI accelerator "Cambricon 1A," which was integrated into Huawei Mate 10, achieving commercial landing of AI chips.
2017: Released three new products (with better energy efficiency than "Cambricon 1A"), covering applications in image recognition, security monitoring, and intelligent driving.
2018: Launched "Cambricon 1H8" and "Cambricon 1H16" for the vision field, "Cambricon 1M" for intelligent driving, and the artificial intelligence system software "Cambricon NeuWare."
2019: Released the "Cambricon MLU220" chip for edge artificial intelligence computing, forming a complete product line of "terminal-edge-cloud" intelligent chips.
Listing and Expansion (2020-Present):
July 2020: Cambricon was listed on the STAR Market (stock code: 688256), becoming the "first AI chip stock in China." The opening price on the first day was 250 yuan per share, a 288% increase from the issue price, and the market value once exceeded 100 billion yuan.
2023: Launched the "Cambricon MLU590" chip (7nm process), whose reasoning energy efficiency was comparable to international giants, supporting mainstream domestic large models and ushering in an outbreak in the cloud product line.
2024: Achieved quarterly profitability, with revenue growing by 43 times year-on-year.
2025: Cambricon's market value once exceeded 600 billion yuan, becoming one of the "leading technology stocks" in the A-share market.
Social Positions:
2023: Elected as a member of the 14th National Committee of the CPPCC and Vice Chairman of the Promotion Committee for Technological Innovation of the Central Committee of Jiusan Society.
2025: Awarded the title of "Fifth National Outstanding Private Entrepreneur in the Construction of Socialism with Chinese Characteristics."
IV. Main Achievements and Honors
Chen Tianshi's achievements cover both academic research and business practice:
1. Academic Achievements
Proposed the concept of "algorithm-architecture-chip" collaborative design, promoting the transformation of AI chips from "general-purpose" to "specialized."
Published multiple top international papers (such as ASPLOS and MICRO Best Papers), leading the academic development in the field of AI chips.
Received the Second Prize of the National Natural Science Award, the Outstanding Achievement Award of the Chinese Academy of Sciences, and other academic honors.
2. Business Achievements
Founded Cambricon, becoming a global pioneer in the field of intelligent chips. The company's valuation once exceeded 600 billion yuan.
Launched multiple independently developed AI chips (such as "Cambricon 1A" and "Cambricon MLU590"), achieving commercial landing and supporting products from mainstream manufacturers such as Huawei and Xiaomi.
Led Cambricon to become a "core domestic enterprise" in "import substitution," forming a complete product line in the fields of cloud, edge, and terminal.
V. Personal Traits and Philosophy
Chen Tianshi's personality and philosophy are centered on three keywords: "focus," "pragmatism," and "innovation":
Focus: Devoted to the field of AI chips for more than a decade, resisting the temptation of short-term profits and adhering to "sticking to technology."
Pragmatism: Emphasized that "chips are silicon lumps and need to support downstream applications," transforming academic achievements into practical products.
Innovation: Proposed the concept of "Cambrian Explosion" in the AI chip industry, believing that AI chips will promote the arrival of the intelligent era and adhering to "doing difficult but correct things."
VI. Summary
Chen Tianshi's path from "prodigy" to "AI chip entrepreneur" embodies the combination of "academic dedication" and "business practice." The Cambricon he founded not only promoted the development of China's AI chip industry but also provided a "Chinese solution" for the global intelligent era. His personal resume, which combines "educational background," "academic achievements," and "business achievements," constitutes a typical sample of a "technology entrepreneur," inspiring more young people to devote themselves to hard technology fields.
What TrendCompass Really Shows — And What It Does NOT🧭 What TrendCompass Really Shows — And What It Does NOT
UIA TrendCompass v1.0 is a real-time market structure interpretation tool.
It is designed to help users understand how market trends evolve — not to predict where price will go next.
This article explains what TrendCompass actually shows, and just as importantly, what it deliberately does not show.
1️⃣ TrendCompass Is About Structure, Not Signals
TrendCompass identifies four recurring market structure states based on price behavior and trend continuity:
T — Trend Start
The initiation of a directional market structure.
E — Trend Extension
Continuation of an existing trend structure.
H — Structural High / Low
Confirmation points within ongoing trend development.
X — Trend Exit / Transition
Structural weakening or transition phase.
These labels are descriptive, not predictive.
They help interpret what the market has structurally done,
not what it is guaranteed to do next.
2️⃣ What TrendCompass Does NOT Do
To avoid any misunderstanding, TrendCompass does not:
❌ Provide buy or sell signals
❌ Predict future price movements
❌ Indicate trade entries or exits
❌ Guarantee accuracy or profitability
TrendCompass does not replace strategy design, execution rules, or risk management.
It is a lens for interpretation, not a decision engine.
3️⃣ Why Structure ≠ Prediction
Market structure is observable only after price has formed it.
This is intentional.
By avoiding prediction, TrendCompass focuses on:
Trend initiation vs. continuation
Structural confirmation vs. transition
Context clarity over outcome guessing
Many trading errors do not come from poor execution,
but from misreading the market’s structural phase.
TrendCompass exists to reduce that specific mistake.
4️⃣ Who This Tool Is Designed For
TrendCompass is intended for:
Users studying price action and market structure
Analysts who prefer contextual understanding over signals
Educational and research-oriented market analysis
It is not designed for:
Signal-following trading
Automated execution systems
Short-term profit optimization
5️⃣ Final Note
All labels generated by TrendCompass are based on historical price data.
They do not forecast future market movements.
This script is provided “as is”, with no guarantee of accuracy or performance.
Users should always combine structural interpretation with their own analysis and risk management framework.
TrendCompass v1.0
A tool for understanding market structure — not predicting the market.
金汇得手:黄金4374受阻回落 今日反弹主空操作 美元指数昨天收长影阳线,今天周五收周线,关注98.2-97.8区间,强阻力98.5,强支撑97.5,大概率跑不出这个大区间。黄金昨天来回震荡洗盘,后半夜大涨4374附近回落,日线收十字星。按照昨天走势,二次上探没新高回落,大概率会有不错的回调。所以操作上我们高空为主,阻力日内高点4337附近,也是日内多空分水,触及不破小止损空。先看4321附近,破位看昨天低点,要想大的下跌空间,必须有效下破4308或4300趋势线支撑。如果行情走震荡,接近4348-4353可以继续空。操作建议:接近日内高点空,震荡的话4348上方空,目标4321破位看4308。接近4300多,有效下破反弹再空。昨天亚盘4330多,欧美盘4322多,4333空,4310多,4375空。原油果然和以前一样,大阳之后不延续,昨天收阴,今天反弹56.3附近空,目标55.1附近,不破小止损多。震荡的话接近昨天高点再空。分析仅供参考,实盘为准。
MELIMercadoLibre 是一个复杂的生态系统,整合了市场、物流、金融科技(Mercado Pago)和数字广告。
公司优势
MELI 的主要竞争优势不在于其各项服务本身,而在于这些服务之间的深度整合,从而为用户打造了一个“闭环”。
物流(Mercado Envíos):其自有物流中心网络,产能同比增长 41%,确保了该地区最快的配送速度。
金融科技(Mercado Pago):Mercado Pago 是一家成熟的数字银行,拥有 7200 万月活跃用户。Mercado Pago 的信用卡已成为巴西平台上使用最广泛的信用卡,超过 50% 的营业额来自外部交易,充分展现了其内在价值。
广告(Mercado Ads):增长最快的业务板块,利用独特的第一方数据进行精准投放。
运营和财务业绩
2025年第三季度营收:74亿美元(同比增长39%)
总支付额(TPV):712亿美元(同比增长41%)
贷款组合:110亿美元(同比增长83%)
管理资产:151亿美元(同比增长89%)
同时,公司保持着投资级信用评级(惠誉评级为BBB-),能够获得低成本融资。
弱点和风险:
信用风险和利润压力
最大的风险来自金融科技领域。贷款组合的快速增长(+83%)超过了营收增长,2025年前九个月的坏账拨备增加了58%。这直接影响了净利润,净利润仅增长13%,远低于营收增长(+37%)。
扣除各项因素后的净息差(NIMAL)下降320个基点至21%。投资者担忧激进的信贷扩张可能在宏观经济下行时引发违约潮,这种担忧不无道理。
战略投资导致息差收窄🔎






















