According to ChainCatcher citing Artificial Analysis Intelligence Index v5, B.AI launched Anthropic Claude Fable 5.2 on November 10, making it available through its API and Web Chat. The model has a 1-million-token context window, supports up to 128,000 output tokens, and keeps adaptive thinking enabled for complex reasoning, software engineering, and long-running agent workflows. At its highest configuration, Claude Fable 5.2 scored 65 on the index, topping the leaderboard and exceeding the median score of 40 for comparable models. Its generation speed is up to 30% faster than Fable 5.1. Pricing is $8 per million input tokens and $40 per million output tokens, with prompt caching supported.

Elon Musk announced plans for the TerraFab semiconductor megafactory in March this year and said in October that it would begin mass production in the second half of 2028. According to Sina Finance, the project has reportedly moved into substantive implementation. Musk plans to build the wafer fab in Texas with Tesla and SpaceX to secure an independent chip supply. Under the plan, it would produce computing chips each year to support a combined one terawatt of computing capacity for AI and robotics businesses. The semiconductor industry has remained skeptical. Building a conventional wafer fab typically takes about three years, and industry observers have questioned whether a company without semiconductor manufacturing experience could reach mass production within that period. A media report on the 9th said TerraFab had recently placed large orders with Japanese semiconductor equipment makers. The project has drawn a new major customer to the industry, while its requests, which reportedly exceed customary industry standards, have puzzled many suppliers. An executive at a Japanese semiconductor equipment company told the media that Elon Musk, rather than Kioxia or Rapidus, was the hottest topic in the semiconductor industry. Sources said TerraFab had informally set a goal of producing 1 million silicon wafers per month from 2032 onward. That would be 40 times the monthly capacity of Rapidus’s plant in Chitose, Hokkaido, which is 25,000 wafers. Of the 1 million wafers, about 700,000 would be used for memory chips and more than 300,000 for logic chips. Intel is also participating in the project and is expected to provide manufacturing process technology during mass production. Several TerraFab procurement executives visited Japan in early November to discuss details for mass production with local suppliers. One equipment maker received an order requiring delivery within a few months. A company representative told the media, “The requirements for this order are unreasonable, but we have to cooperate as much as possible.” The unprecedented delivery schedule has caused a stir across the industry. Advantest CEO Douglas Lefever recently said at an earnings conference that TerraFab was moving from concept to implementation and could become a significant business. TSMC President C.C. Wei said last month that the two sides were discussing how to use TerraFab’s capacity. Reports said TSMC was considering leasing part of the site to build its own wafer fab, but Musk said his company would operate the factory itself. The mass-production plan remains uncertain, as does whether suppliers can meet Musk’s demands. Market concerns also include possible delays to approvals for related AI investment projects following a change in the Texas governorship after the U.S. midterm elections. If realized, the plan for high-performance AI chips would make TerraFab a new competitor to South Korean semiconductor giants such as Samsung and SK Hynix.

According to Jin10, WTI crude futures fell $2.01 intraday to $84.84 per barrel, down 2.31%, while Brent crude futures fell 2% to $90.66 per barrel.

Nvidia is facing scrutiny over a US chip-smuggling probe as AI stocks rebound, according to Bloomberg.

NVIDIA (NASDAQ: NVDA) is scheduled to report its third-quarter fiscal 2027 results on November 17, 2026, after the market close, with a conference call at 5:00 p.m. ET. Investors are watching whether strong demand for artificial intelligence infrastructure can sustain the chipmaker’s rapid growth. Expectations remain high as major technology companies continue investing heavily in AI data centers, putting NVIDIA’s revenue growth, data center performance, and forward guidance under close scrutiny. Consensus Estimates & Expectations Revenue Estimate: According to analyst estimates compiled by Charles Schwab as of November 4, NVIDIA is expected to report revenue of approximately $109.2 billion, up 91.5% from $57.01 billion a year earlier. Broader consensus figures cluster in a similar range of roughly $108.8–$109.2 billion. This sits near the upper end of the company’s own guidance of $108.0 billion, plus or minus 2% (approximately $105.8–$110.2 billion). The guide assumes no data-center compute revenue from China. Hitting the midpoint would represent roughly 12% sequential growth from the prior quarter’s $96.2 billion. Earnings Per Share (EPS): Adjusted earnings per share are forecast at $2.48, representing year-over-year growth of 90.5% from $1.30 in the year-ago quarter. Other compilations place the consensus near $2.47–$2.49. In the second quarter of fiscal 2027, NVIDIA delivered adjusted EPS of $2.22, beating expectations. Data Center & Blackwell Demand: Data Center remains the dominant driver and is expected to account for the vast majority of revenue. In the prior quarter it reached a record $89.0 billion, up 117% year over year and more than 92% of total sales. Blackwell Ultra continued to ramp strongly, particularly with hyperscalers. The next-generation Vera Rubin platform began production shipments in August 2026 and is expected to contribute roughly 20% of Data Center revenue in the third quarter. Management has described purchase orders from major hyperscalers, AI clouds, and system makers, with deployments already underway at partners including CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure, and others. Hyperscale customers and the AI Clouds, Industrial, and Enterprise (ACIE) segment both showed robust sequential growth in the prior period. Broader industry data points to continued heavy capital spending by the largest cloud providers, supporting the near-term demand picture even as supply constraints persist. Analyst Views on the Stock Wall Street remains overwhelmingly bullish on NVIDIA despite recent share-price weakness and governance questions. As of early November 2026, the consensus rating is Strong Buy or Buy, with roughly 50–60 analysts covering the stock and only a handful of Holds and essentially no Sell ratings. Average 12-month price targets cluster in the $320–$340 range, implying substantial upside from recent trading levels near $178–180. Notable targets include Bernstein’s Stacy Rasgon at $400 (maintained Outperform), Bank of America at $350, RBC Capital at $330, Citi at $315, and Goldman Sachs and Morgan Stanley near $300. Most analysts argue that underlying AI infrastructure demand and product cycles (Blackwell Ultra and the Vera Rubin ramp) remain intact and that any near-term distraction is unlikely to derail shipments or the multi-year growth trajectory. A small minority have turned more cautious, with at least one firm downgrading to Hold citing potential management turmoil, but the overall stance is that the stock’s risk-reward has become more attractive after the pullback. Key Headwinds & Market Overhang Several factors are weighing on sentiment heading into the print. Gross margins are guided to approximately 74% (plus or minus 50 basis points), down from 75% in the second quarter, with management signaling further pressure in the fourth quarter (potentially bottoming near 71–72%) due to rising memory and component costs. Supply remains a constraint rather than demand. Long-term supply and capacity commitments rose sharply in the prior quarter, and the company continues to describe itself as supply-limited even while guiding for strong sequential growth. Geopolitical and regulatory issues form a persistent overhang. The outlook explicitly assumes zero China data-center compute revenue amid ongoing export controls. Broader political and policy uncertainty has contributed to recent stock weakness. Recent reporting has also drawn attention to personal and governance matters involving CEO Jensen Huang. Major news outlets have reported that Huang admitted to an extramarital affair with Alice Huang, a Megaspeed executive whom U.S. intelligence concluded was a Chinese intelligence asset. NVIDIA’s board launched an independent investigation into whether he was the target of a foreign intelligence effort, while a shareholder lawsuit alleges inadequate disclosure of related risks. The board has expressed confidence in Huang’s leadership and stated it has no reason to believe national security or company secrets were compromised; Huang has said he supports the probe. Analysts have noted that the episode heightens long-standing key-man risk concerns given Huang’s central role and the absence of a clear successor, though most still see it as a temporary overhang rather than a fundamental demand problem. NVIDIA shares have declined substantially from recent highs in the weeks leading up to the report, reflecting a combination of these concerns and high expectations. While the company has a strong track record of beating its own guidance and analyst estimates, the bar is elevated, and forward commentary on Vera Rubin ramp, margins, supply, and the durability of AI infrastructure spending will likely determine the market’s reaction. What to Expect in NVIDIA’s Earnings Report Revenue Growth: Revenue is expected to reach approximately $109.2 billion, up 91.5% year over year. Earnings Performance: Adjusted EPS is forecast at $2.48, nearly doubling from a year earlier. Data Center Demand: Investors will look for continued growth in AI chip sales, particularly Blackwell Ultra deployments. Vera Rubin Ramp: The market will assess early contributions from the next-generation platform and its shipment outlook. Gross Margins: Rising memory and component costs could put pressure on margins, with further declines expected in the coming quarter. China Exposure: Investors will watch for any updates on export restrictions and the outlook for data center sales in China. Forward Guidance: Management’s commentary on supply capacity, hyperscaler spending, and AI infrastructure demand will be critical to the stock’s reaction.

Initia coordinated with validators to pause its L1 at 14:18 (UTC+8) on November 10 after detecting unauthorized transfers from a custodial account. According to Odaily, an initial investigation indicated that an attacker used compromised Rollup infrastructure keys to forge IBC transfers. A preliminary review found that about $605,000 in USDC and another $10,000 worth of tokens had left Initia through Noble before the pause. Most of the other assets transferred from the custodial account remained on the paused L1.

Galxe宣布关停任务平台、团队并入Gravity，3600万用户的撸毛时代落幕，转战L1的增长故事仍需市场检验。

仅支付15美元的网络费用，黑客就诱骗币安工程师安装了恶意代码，从而控制了一个热钱包，并在一次交易中盗走了约25亿美元。即使币安做出反应后，该访问权限仍然在区块链上保持活跃，最终被盗金额高达63亿美元。

英伟达,加密货币,资产,法律,英伟达大瓜刷屏 真与科技大佬离婚 加密货币该怎么分？ 金色财经,加密货币离婚分割司法要点分析。

当AI巨头创始人的个人声誉与公司市值、产业链信心深度绑定，危机公关到底能做什么，又不能替代什么？

AI Agent,DeepSeek,Harness,LLM,Token,Harness大战 金色财经,AI时代，必争执行层。

超过48亿美元的AI设备采购，一家公司的资产负债表从3300万美元扩大到约30亿美元。

如果地缘风险缓和、利率预期改善，风险偏好逐渐恢复，加密市场也有机会获得新的支撑。

Anthropic,算力,Claude,LLM,AI Agent,IPO前夜 Anthropic再亮两张“王牌” 金色财经,1000个AI智能体，30小时攻克费马大定理。

SEC,美股,代币化股票,纽交所,沉默即同意：美股代币化的艰难之路 金色财经,为何美国企业界不愿任由第三方将自家股票代币化。

朝鲜黑客利用其在币安（Binance）窃取的63亿美元资金，制造了六种毫无价值的“迷因币”，并用它们转移了约43万美元的赃款，使之看起来像是正常的交易利润。区块链公司Elliptic追踪到了这一活动，其手法与去年Bybit被黑客攻击后黑客使用的手段如出一辙。