Core Viewpoint - The "Magnificent Seven" stocks have shown mixed performance in 2025, with varying potential for 2026, leading to a ranking of these stocks from avoid to strong buys [1]. Group 1: Stock Rankings and Performance - Nvidia is ranked as the top stock, expected to continue strong growth due to its position as a primary supplier for AI computing, with anticipated capital expenditures in data centers rising significantly [21][22]. - Alphabet follows, having experienced over 60% growth in 2025, driven by advancements in AI and a strong core business in Google Search, positioning it well for 2026 [17][20]. - Amazon, despite a poor performance in 2025 with only a 3% stock rise, is expected to rebound in 2026, particularly due to growth in its AWS and advertising services [15][16]. - Meta Platforms had a strong 2025 until Q3 earnings, with a 26% revenue increase attributed to AI, but concerns over capital expenditures may impact its performance [13][14]. - Microsoft is positioned for impressive growth in 2026, benefiting from investments in OpenAI and its cloud computing services, with a 14% increase in stock value in 2025 [11][12]. - Tesla is ranked lower due to challenges from the end of EV tax credits and shrinking margins, suggesting it may be wise to avoid this stock in 2026 [7][10]. - Apple ranks last, with stagnant revenue growth and a high stock price relative to earnings, indicating underperformance potential in 2026 [3][6]. Group 2: Financial Metrics and Market Insights - Nvidia's GPUs are in high demand, with expectations of record-setting capital expenditures in the AI sector, indicating a strong market position [21][22]. - Alphabet's market cap stands at $3.7 trillion, with a gross margin of 59.18%, reflecting its robust financial health [19][20]. - Amazon's market cap is $2.4 trillion, with AWS growth at 20% and advertising services at 24% growth in Q4, highlighting its operational strengths [16]. - Microsoft has a market cap of $3.6 trillion and a gross margin of 68.76%, showcasing its profitability and growth potential [12]. - Meta's revenue growth of 26% in 2025 demonstrates its ability to leverage AI, although future capital expenditures remain a concern [14]. - Tesla's margins have declined, impacting its earnings per share, which suggests a need for strategic adjustments [7][10].

Core Viewpoint - Apple stock (AAPL) is considered a long-term buy through dollar averaging, driven by three specific catalysts: the iPhone 17 supercycle, the monetization of Apple Intelligence, and a shift towards high-margin Services [1] Group 1: Catalysts for Investment - The iPhone 17 supercycle is expected to significantly boost sales and market presence [1] - The monetization of Apple Intelligence indicates a strategic shift towards higher-margin services, enhancing overall profitability [1] - The transition to high-margin Services is anticipated to provide a stable revenue stream, further supporting the investment thesis [1]

机构：光大证券 研究员：付天姿/黄铮 事件: 12 月17 日，小米集团举办2025 年"人车家全生态"合作伙伴大会，核心亮点集中于大模型技术突 破、全生态协同深化与开放生态扩容。 Mimo-V2-Flash 大模型：速度+成本双优，期待未来的硬件融合与商业化进展。小米发布新一代基座大 模型Mimo-V2-Flash，1)性能上，该模型以"309B 总参数 + 15B 激活参数" 的轻量化设计，在代码能力与 Agent 能力的全球开源模型评测中位列 Top2，核心评测指标超越参数量2~3 倍的DeepSeek V3.2 与Kimi K2-Thinking，技术实力行业领先。2)技术架构上，Mimo-V2-Flash 采用5:1 比例的Hybrid SWA（滑动窗 口注意力+全注意力）架构，固定KV Cache 适配主流推理框架，并搭配3 层MTP 推理优化，实现 2.0~2.6 倍推理提速，单机吞吐量5000-15000 token/秒，单请求吞吐量115-150 token/秒。3)推理效率上, Mimo-V2-Flash 成本与速度优势显著，其推理速度是推理成本相似的DeepSeek V3.2 的3 倍， ...

【#常规操作可能无法彻底清除手机信息#】#恢复出厂设置能清除手机信息吗#近年来，随着电子产 品"以旧换新"和二手交易的快速扩张，大量旧手机、旧电脑进入回收渠道，但是这些电子产品中存储的 用户照片、通讯录等个人信息如果没有被有效清除，就容易引发消费者的隐私泄露。技术人员介绍，消 费者在处置旧电子产品时，仅采用"删除文件""恢复出厂设置"等常规操作实际上很多时候并没有彻底清 除用户信息，依靠一些技术手段就能恢复被删除的数据。@央视网 #央视曝光旧手机回收猫腻# ...

因此，"小米17Ultra一定会涨价，而且我觉得还要涨得有点多。但是跟内存成本上涨（比起来），我觉得还是低的。" 据报道，小米 17 Ultra 搭载的一英寸主摄和 2 亿像素长焦镜头也已曝光，其中主摄搭载第三代 LOFIC 技术，动态范围相比上代提升十几倍；而长焦镜头的 体积比 15 Ultra 大 35%，成本大约是 15 Ultra 的两倍，有可能是行业最复杂的长焦结构。 此前有爆料称，小米 17 Ultra 手机开发代号为"哪吒"（Nezha），延续了小米 15 Ultra 标志性的圆形相机模组设计，采用高通第五代骁龙 8 至尊版芯片，预 装澎湃 HyperOS 3.0，此外已确认其国际版机型将支持卫星通信功能，为用户在偏远地区提供可靠的连接保障。 卢伟冰表示，从2022年底至今差不多三年间，AI迎来了爆发式的增长。根据整体判断，2025、2026、2027三年都会是内存成本上涨点。内存价格的猛涨， 进而会带来手机成本的大幅上升。 此前小米15Ultra发布时，官方曾称这是"最后一次6499元"。卢伟冰直言，当时并没有完全考虑到内存，仅仅是基于处理器成本、相机配置上涨而作出的判 断。而17Ultra更 ...

来源：环球网 【环球网科技综合报道】12月21日消息，据《福布斯》报道，苹果公司正在研发其首款可折叠iPhone已 是公开的秘密。然而最新报告显示，这款备受期待、原定于明年秋季亮相的产品，可能正遭遇生产方面 的挑战。 不过先别急着期待——情况可能更为复杂。郭明錤补充说明："由于早期生产良率及产能爬坡面临挑 战，实际顺利出货可能要等到2027年。加之预期需求强劲而供应有限，可折叠iPhone在2026年底前很可 能持续面临缺货局面。" 苹果近年来已很少遇到新产品严重供不应求的情况。虽然2017年iPhone X曾将发售时间推迟数周，但上 一次出现类似程度的交付延迟，还要追溯到2016年第一代AirPods发布之时。 那么这意味着什么？先从时间线看起。下一代iPhone旗舰机型预计将于2026年秋季发布——尽管iPhone 17e可能提前至春季上市。目前距离发布还有近十个月，这段时间通常足以让苹果调整生产节奏。 这份报告来自备受业界关注的天风证券分析师郭明錤。他在最新研究报告中指出："可折叠iPhone的研 发进度目前落后于预期，但预计仍能在2026年下半年正式发布。" 回顾2025年初，iPhone Air也曾 ...

Core Concept - TORRAS announces its first global artist collaboration with Gawx, introducing the GUARDIAN collection that emphasizes the theme of "guardianship" in creative expression [1][2] Product Details - The GUARDIAN collection features the Ostand Q3 Air phone case, designed to be a creative canvas that combines artistic expression with military-grade protection [4][9] - The collection includes original artwork by Gawx, collectible stickers, and a 360° magnetic Ostand, enhancing the user experience by merging art with functionality [9] Artistic Collaboration - Gawx, a recognized visual artist and filmmaker, brings a unique emotional and imaginative perspective to the collaboration, having won the 2025 IMI Award and being named RØDE Creator of the Year 2024 [3] - The collaboration is centered around the concept of DUALITY, representing two contrasting creative lives inspired by Day and Night [4][5] Thematic Elements - The Day theme portrays a calm boy and his dog capturing subtle inspirations in a serene morning setting, while the Night theme features a confident girl and her dog seizing vibrant moments in a chaotic urban environment [6][7] - Both narratives emphasize the importance of protection for creativity, with TORRAS acting as a guardian for the creators' devices and ideas [8] Launch and Promotion - The GUARDIAN collection is set to launch on December 20th, with a retail price of $79.99 and a promotional offer of 10% off for the first five days [11] - Additional winter-themed designs are introduced, including a Christmas Edition, enhancing the seasonal appeal of the product lineup [10][12]

Group 1 - Apple Inc. is recognized as one of the 14 best S&P 500 stocks to buy, with JPMorgan maintaining an Overweight rating and a price target of $305 [1] - Citi analyst Atif Malik raised the price target for Apple to $330 from $315, citing expected increases in iPhone sales as users upgrade from older models [2] - CLSA also increased its price target for Apple to $330 from $265, highlighting a surge in iPhone sales that is anticipated to drive revenue and earnings growth in FY26-27 [3] Group 2 - Wall Street analysts currently have a Moderate Buy rating on Apple, with an average one-year price target of $298.72, indicating a potential upside of 9% [3] - Apple plans to expand its iPhone product line to include seven variants by 2027, reflecting its commitment to innovation in consumer electronics [4]

Vivo X300 Pro Price and Availability The X300 series was officially launched in India on December 2, 2025 and comes in a single configuration 16 GB RAM + 512 GB storage. It is available in Elite Black and Dune Gold colour options with a retail price of Rs 1,09,999. Pre-bookings for the X300 Pro started soon after launch, and sales commenced on December 10, 2025, via official retail channels, including the Vivo India website, major e-commerce platforms, and offline partner stores.Design and In-Hand Feel If y ...

过去一年时间里，手机厂商在AI方面的动作越来越密集，荣耀YOYO接入更多第三方智能体，实现系统层与应用层的AI能力对接；华为小艺只需一句指令便 能在应用间穿梭，帮你完成一些任务。 尽管这些AI能力愈发强大，但如果把这些功能拆开来看，会发现一个很现实的情况，本质上，这些AI依然需要联网才能使用。也就是说，手机AI仍困在端 云协作的阶段，还未进一步。 近期，X平台Semi-retired-ing透露，三星将在即将发布的Galaxy S26系列上准备一套能够在本地运行的大模型，用于实现大部分AI功能。这套大模型甚至拥 有高级权限，能在必要时清除内存，留出更多空间，以确保随时响应用户的需求。 （图源：Oneleaks） 实际上，三星在2023年就曾展示过一套名为"高斯"的本地大模型，也曾被指出这套模型已经在Galaxy S25系列里预装。但不知出于何种原因，三星一直在力 推Google的Gemini，几乎未再提及"高斯"。直到近期，三星的本地大模型才被重新提及。 在大多数厂商仍然以云端为主的阶段，三星为何要尝试把模型真正放进手机里？是希望通过这种方式"弯道超车"？还是移动端已经拥有了本地部署大模型的 能力？无论答案是 ...