免責聲明更新： 大叔美股筆記歡迎各位投資同好共襄盛舉，為維護個人權益，在瀏覽各個投資筆記前，請先閱覽以下聲明，閱覽者將視為同意本聲明條款： 1. 大叔美股筆記對於所發表之投資筆記、分析數據及參考資料內容，其立論依據皆係合法引用自網路上公開資訊以及合法管道取得之資料，不涉及任何抄襲等侵害智慧財產權之行為;對於閱覽者將大叔美股筆記所發表之投資筆記，在未經同意下，從事行銷推廣、背書代言、營利等商業相關活動，無論是否獲利，大叔美股筆記將採取法律行動追究其責。 2. 大叔美股筆記所發表之投資筆記、分析數據及參考資料內容，不得作為任何交易之依據，閱覽者依筆記或數據資料內容進行任何投資行為所產生之風險及盈虧，未經自行了解判斷，而導致交易損失風險，應自行負擔，大叔美股筆記不對使用者之投資決策負任何責任。 3. 大叔美股筆記所發表之投資筆記對於所提及之其他個人、公司或組織，大叔美股筆記均不對其交易安全、商品及服務內容負任何擔保責任，其間之所有交易、擔保、賠償責任及售後服務皆係存在於閱覽者與商品或勞務提供者之間，閱覽者如因商品或服務提供者發生任何變故或糾紛，大叔美股筆記不負任何法律責任。 4. 大叔美股筆記中，涉及到其他個人、公司或組織之網站連結，對被連結之網站或其所提供之任何產品、服務或資訊，大叔美股筆記不擔保其正確性、真實性、即時性、及完整性。 5. 閱覽者同意由大叔美股筆記保留修改本聲明之權利，如有變更，將公佈於大叔美股筆記社群平台上，不另行通知。 Disclaimer Updates: Uncle Stock Notes welcomes all investment enthusiasts to participate. To safeguard your rights and interests, please read the following disclaimer carefully before reviewing any investment notes. By accessing or reading the content, you are deemed to have agreed to the terms set forth herein: 1. All investment notes, analytical data, and reference materials published by Uncle Stock Notes are based on legally sourced information, including publicly available online data and materials obtained through lawful channels. No infringement of intellectual property rights, including plagiarism, is involved. Any unauthorized use of Uncle Stock Notes’ content for marketing, promotion, endorsement, commercial gain, or related activities—whether for profit or not—without prior consent is strictly prohibited. Uncle Stock Notes reserves the right to take legal action against any such violations. 2. The investment notes, analytical data, and reference materials provided by Uncle Stock Notes are for informational purposes only and shall not be construed as a basis for any trading decisions. Any investment actions taken by readers based on such content are undertaken at their own risk. Readers are solely responsible for any profits or losses incurred, especially where decisions are made without independent evaluation and judgment. Uncle Stock Notes assumes no responsibility or liability for any investment decisions made by users. 3. With respect to any individuals, companies, or organizations mentioned in the investment notes, Uncle Stock Notes makes no representations or warranties regarding the security of transactions or the quality of any products or services provided. All transactions, guarantees, compensation, and after-sales services are solely between the reader and the respective provider. Uncle Stock Notes shall bear no legal liability for any disputes, losses, or issues arising between readers and such third parties. 4. Any external links to websites of individuals, companies, or organizations provided within Uncle Stock Notes are for reference only. Uncle Stock Notes makes no guarantees regarding the accuracy, authenticity, timeliness, or completeness of the content, products, or services offered on such linked websites. 5. Uncle Stock Notes reserves the right to amend this disclaimer at any time. Any changes will be published on its official platforms without prior notice.

艾司摩爾 $ASML 2026 年第二季財報分析 $ASML 這一季最亮眼的是客戶寧願花錢升級手上的舊機台，只為了多擠出一點產能。這個訊號，比任何營收數字都更能說明 AI 晶片的產能有多緊。 同一時間，ASML 把全年營收指引拉到 €430–450 億，半年內上修約兩成；2027 年的 EUV 產能也已接近被訂單填滿。 今年有很大一部分業績壓在第四季，股價也早已先走一步。這麼好的公司，現在這個價格還剩多少安全邊際？大叔這篇一次拆給你看。 文末附錄：大叔把 ASML、應用材料（AMAT）、科林研發（LRCX）、東京威力科創（TEL）和 Onto Innovation（ONTO）放進同一張 AI 資本支出版圖，比較誰的護城河最深、誰對記憶體最敏感、誰的估值最便宜。 unclestocknotes.substack.com…

中秋佳節快樂，闔家平安。 謝謝大家這麼多年來對小弟、對「大叔美股筆記」的支持與陪伴。 投資的路很長，市場每天都有新的故事，我們就一起學習、一起往前走。 到了中秋這個日子，股票可以晚一點看，財報也可以晚一點研究。 有時間，就多陪陪家人。 畢竟我們努力投資、努力工作，最後想守護的，還是身邊最重要的人。 祝大家中秋佳節愉快，也祝福大家和家人都健康、平安、開心。 大叔敬上

$ORCL Oracle 對 Project Jupiter 發出不可抗力通知，市場以為它想卸責。但合約攤開來看，電力是它的責任、租約不能終止、債務利息也還是它付。這份通知能改變的只有時間，改變不了風險歸屬。 unclestocknotes.substack.com…

At Connect, $Meta introduced Charm, a pocket-sized AI device powered by its Muse agent. The device is roughly the size of an AirPods case, features a 2-inch touchscreen and 5G connectivity, and is expected to ship during the year-end holiday season. Pricing has not yet been announced. Muse is also coming to Meta’s smart glasses. The company additionally introduced $1,299 VR Glasses and said it expects to offer more than 100 AI eyewear models by the end of the year. The bigger strategic move is clear, Meta is trying to extend Muse from a popular app into an always-on hardware interface. That could give the company a more direct way to control how users interact with AI agents, commerce, and advertising throughout the day. But the real investment questions are still much more basic than the launch event itself. What matters is: Hardware usage Subscription conversion Gross margin Those metrics will tell us far more than the presentation lights. The main risks remain: • Privacy concerns • Battery life • Pricing that may be too high • Limited consumer interest in carrying yet another dedicated device 🧔 My Take Meta is clearly trying to move Muse beyond software and make it part of the user’s daily physical environment. If that works, the opportunity is much larger than just another AI app. It could become a new distribution layer for AI Agents → Commerce → Advertising → Hardware But hardware is unforgiving. The real proof will come from usage, retention, monetization, and margins — not from how impressive the demo looks on stage.

Meta在Connect發表搭載Muse代理的Charm隨身AI裝置，體積約如AirPods盒、配備2吋觸控螢幕與5G，預計年底假期出貨但價格未公布。Muse也將進入 $Meta 智慧眼鏡，公司另推出1,299美元VR Glasses，並預計年底前提供超過100種AI眼鏡款式。 Meta正把Muse由熱門App延伸為全天候硬體入口，希望掌握代理、商務與廣告互動；但硬體使用率、訂閱轉換及毛利仍比發表會燈光更重要。 主要風險： 隱私疑慮、電池續航、價格過高及消費者對另一件隨身裝置興趣有限。

September 24 Market Close The S&P 500 slipped 0.02%, the Nasdaq gained 0.01%, and the Nasdaq 100 rose 0.03%. The Dow fell 0.31%, while the Russell 2000 declined 0.11%. Treasury yields moved higher again. The 2-year yield rose to 4.87%, the 10-year climbed to 5.18%, and the 2s10s spread widened from 26 basis points to 31 basis points. The 30-year yield rose to 5.47%. This is a classic bear steepening move led by the long end of the curve. Oil also moved sharply higher. WTI gained 2.66% to $94.61, while Brent rose 3.41% to $106.60. The global crude-oil risk premium continues to build. The VIX, however, only rose to 15.67, which still does not indicate a full-blown fear regime. Yesterday’s selloff was driven by the combination of rebounding oil prices and higher Treasury yields. Today, the major indexes stabilized, but the underlying picture continued to deteriorate. Long-term yields moved higher, Brent kept rising, and market breadth remained weak. That makes the case for selective defense even stronger. The focus should increasingly be on filtering out companies with: High financing costs Negative free cash flow Long-duration valuations The VIX is still relatively low, so I would not characterize this as a full risk-off environment yet. For now, the next key levels are: 10-year Treasury: 5.20% Brent crude: $110 VIX: 19 If all three begin moving through those levels together, that would be a stronger signal that the market is transitioning from selective pressure into a broader contraction in risk appetite. On the AI side, I am watching Akamai’s service ramp, contract gross margins, and financing structure, as well as whether Oracle can keep Project Jupiter on track for its 2028 timeline. 🧔 My Take The indexes may look calm, but the pressure underneath the surface is still building. For now, this remains a market where stock selection matters more than simply owning the index. The biggest risks are still coming from the same place: Higher energy costs + Higher long-term yields + Higher cost of capital Until those pressures begin to ease, I would continue to favor businesses with stronger balance sheets, positive free cash flow, and earnings that can support their valuations.

9月24日收盤： S&P 500跌0.02%、Nasdaq漲0.01%、Nasdaq 100漲0.03%；道瓊跌0.31%，Russell 2000跌0.11%。 2年債殖利率升至 4.87%，10年債升至 5.18%，2s10s曲線由26擴大至31個基點；30年債升至 5.47%。這是長端領跌的熊市陡峭化。 WTI漲2.66%至 $94.61，Brent漲3.41%至 $106.60；全球原油風險溢價繼續上升。VIX僅升至 15.67，仍未進入恐慌區。 昨天是油價與殖利率反彈造成的全面回檔；今天指數雖然穩住，但長端利率、Brent與市場廣度繼續惡化。這讓「選擇性防守」更確立，尤其要增加對融資成本、負自由現金流及長久期估值的篩選。VIX仍低，因此尚未升級為全面Risk-off。 接下來看10年債是否突破 5.20%、Brent是否接近 $110；若再加上VIX突破19，才升級為全面風險收縮。AI方面則看Akamai的服務啟動、合約毛利與融資安排，以及Oracle能否守住Project Jupiter的2028時程。

Anthropic has committed to purchase at least $11.6 billion of cloud services from $AKAM Akamai over the next seven years, with an option to expand the relationship by another $9 billion, bringing the potential total to roughly $20.6 billion. A simple average would imply: $11.6B ÷ 7 ≈ $1.66B per year Of course, actual revenue recognition will not be evenly distributed, and Akamai has already said that the agreement does not change its 2026 revenue guidance. But relative to Akamai’s current scale, this is still a very large contract. Akamai generated roughly $1.10 billion of revenue in Q2 2026, while its full-year revenue guidance is around $4.45–4.53 billion. So an $11.6 billion contracted commitment is clearly meaningful for $AKAM. The more interesting point is what Anthropic is actually buying. The agreement is aimed at “accelerating CPU workload demands.” Frontier AI systems are not just GPUs doing matrix multiplication. There is a much broader infrastructure stack behind every model: CPU compute Agent orchestration Data preprocessing Storage and retrieval Inference services Networking Security Memory So the better question for AI infrastructure may be: For every $1 of accelerator compute deployed, how much additional CPU, memory, storage, networking, and power is required around it? That is the real AI infrastructure multiplier. 1. Memory is spreading beyond HBM Akamai expects to invest about $5.5 billion of CapEx to support the initial $11.6 billion commitment. In 2026 alone, the company expects roughly $1.7 billion of incremental CapEx to secure critical supply-chain components. And Akamai specifically mentioned memory. The AI infrastructure memory hierarchy increasingly looks like this: HBM ↓ Server DRAM ↓ Enterprise SSD / NAND ↓ Storage The important point is that AI-driven memory demand is no longer limited to accelerator memory. It is spreading across the entire AI infrastructure stack. For Micron, SK hynix, and Samsung, the AI memory opportunity is therefore broader than HBM alone. 2. Anthropic is moving toward Multi-Cloud + Multi-Architecture This may be one of the most important parts of the story. Earlier this year, Anthropic announced an expansion of its Google/Broadcom TPU capacity. The company now uses multiple compute architectures, including: AWS Trainium Google TPU NVIDIA GPU Amazon remains Anthropic’s primary cloud provider and training partner. Now Akamai is being added for CPU infrastructure. Anthropic’s infrastructure strategy is starting to look increasingly workload-specific: Different workloads are assigned to the compute architecture that offers the best combination of performance, cost, and available capacity. The future AI stack may therefore not be dominated by a single architecture. It may be GPU + TPU + Trainium + CPU + Memory + Networking + Storage. all operating together. 3. Anthropic also becomes a potential Akamai shareholder As part of the agreement, Akamai granted Anthropic warrants equivalent to approximately 7.7 million shares, with an exercise price of $111.33. At full vesting, the warrants could represent roughly 5% of Akamai’s shares outstanding. Approximately 2% is expected to vest alongside the current $11.6 billion commitment. The remaining roughly 3% is tied to the additional $9 billion expansion. In simple terms: Every additional ~$3 billion of cloud purchases could unlock roughly another 1% of Akamai equity. That makes this relationship more than a normal vendor contract. It increasingly looks like Customer + Strategic Partner + Potential Shareholder. If Anthropic shifts more workloads onto Akamai: Akamai revenue rises And if Akamai receives a higher valuation multiple as a result: Anthropic’s warrant value rises, The economic incentives are increasingly aligned. 4. The real question is ROIC There is one important caveat. Akamai is expected to spend roughly $5.5 billion of CapEx to support the initial contract. So the investment case should not stop at $11.6B of contracted revenue. It needs to go further: Revenue → Gross Profit → EBITDA → Free Cash Flow → ROIC The math is important: Contract commitment: $11.6B Associated CapEx: ~$5.5B That means CapEx is equivalent to roughly 47% of the initial commitment. This is a very capital-intensive infrastructure buildout. So the real question is not simply how large the contract is. It is how much economic profit Akamai can generate after paying for the infrastructure required to deliver it. 🧔 My Take This is why I keep saying that the second phase of the AI infrastructure story is not only about who builds the fastest GPU. The more interesting question is: How much infrastructure has to exist around the GPU? CPU. Memory. Storage. Networking. Power. Cooling. Security. The Anthropic × Akamai $11.6 billion agreement is one of the clearest examples yet of how AI infrastructure demand is spreading beyond accelerators and into the rest of the stack. That, in my view, is where the next layer of AI infrastructure research becomes very interesting.

Anthropic 承諾未來 7 年至少向 $AKAM Akamai 採購 $11.6B cloud services，另外還可以擴張最多 $9B，因此潛在總額接近 $20.6B。 簡單平均一下：$11.6B ÷ 7 ≈ $1.66B/year 當然實際 revenue recognition 不會平均分布，而且 Akamai 表示這筆交易不影響 2026 revenue guidance。 拿 Akamai 目前規模來比較，公司 Q2 2026 單季 revenue 約 $1.10B，全年 guidance 約 $4.45–4.53B。$11.6B backlog 對 AKAM 絕對不是小訂單。 Anthropic 這次跟 Akamai 簽的是“accelerating CPU workload demands”。 Frontier AI system 並不是只有GPU 跑 matrix multiplication，還有大量： CPU compute Agent orchestration Data preprocessing Storage / retrieval Inference services Networking Security Memory 現在可以看每增加 $1 的 accelerator compute，會拉動多少 CPU、Memory、Storage、Networking、Power？ 這才是整個 AI infrastructure multiplier。 第一：Memory，Akamai 預計為最初 $11.6B commitment 投入大約$5.5B CapEx，其中光 2026 年就會額外增加約 $1.7B CapEx 用途是提前取得 critical supply-chain components，官方還特別點名including memory。 AI infrastructure memory hierarchy 是： HBM ↓ Server DRAM ↓ Enterprise SSD / NAND ↓ Storage AI memory demand 正從 accelerator memory 往整個 AI infrastructure stack 擴散。這對 MU / SK hynix / Samsung 的意義不只是HBM。 第二：Anthropic 正在 Multi-Cloud + Multi-Architecture 這也是我覺得很重要的一點。 Anthropic 今年稍早已經宣布擴大 Google/Broadcom TPU capacity，而且官方說它同時使用： AWS Trainium Google TPU NVIDIA GPU Amazon 仍是主要 cloud provider 與 training partner。Anthropic，現在又加入Akamai → CPU Infrastructure，所以 Anthropic 的 infrastructure architecture 越來越像： 不同 workload，丟給最適合、最便宜、最有供應能力的 compute architecture。 而是GPU + TPU + Trainium + CPU + Memory + Networking + Storage一起存在。 最後，Akamai 給 Anthropic：7.7M shares equivalent exercise price：$111.33 最多相當於 Akamai 約：5% shares outstanding 其中大約 2% 隨現在的 $11.6B commitment vest；剩下約 3% 則與額外 $9B expansion 掛鉤，而且每增加 $3B cloud purchases → 約再 vest 1%。 這關係比較像： Customer + Strategic Partner + Potential Shareholder，Anthropic 如果把更多 workload 放到 Akamai：Akamai Revenue ↑ 如果 Akamai 因此 rerating：Anthropic warrant value ↑ 等於雙方利益綁在一起。 因為 Akamai 為了拿這張單，自己要投入~$5.5B CapEx，所以重要的是$11.6B Revenue → Gross Profit → EBITDA → FCF → ROIC。 Contract commitment = $11.6B Associated CapEx ≈ $5.5B CapEx 相當於 commitment 的約 47%，這是一個相當 capital-intensive 的 buildout。 這也是為什麼大叔最近一直說，AI 第二階段值得研究的，不只是誰做最快的 GPU，而是「GPU 旁邊還需要多少東西」。今天 Anthropic × Akamai 這張 $11.6B 合約，我認為就是非常漂亮的一個例子。

摩根大通將CoreWeave $CRWV 的評級上調至“增持”，目標價為125美元。 定價走強，7月的費率上漲了25%，短期計算價格約為長期合約價格的3倍，近期簽署的交易價格接近4,000萬美元/兆瓦。 摩根大通表示，定價和利潤率的提升足以抵消不斷增長的債務。

This is the autopen... it's Biden." 🤣 President Trump shows off the legendary Presidential Walk of Fame to President Xi.

Chinese President Xi said Make America Great Again.

長江存儲在慕尼黑拿下第一回合 德國慕尼黑一紙判決,讓「中國告贏美光」洗了一輪版。 但把專利機制拆開來看，兩項未經實質審查的實用新型、一審判決、美光已上訴、效力只在德國，而且不涉及 HBM。對美光近期財務的衝擊，比標題小得多。 只是看到媒體標題有點過頭，出來做個平衡報導。 就那麼剛好，近期美光要財報了。 $MU unclestocknotes.substack.com…

美國8月非農就業增加 16.2萬人，失業率維持4.1%；相較過去12個月平均每月僅增加3.1萬人，最新數據明顯較強。 平均時薪月增0.3%、年增3.1%。餐飲業增加5.9萬人，但資訊業減少2.3萬人。經濟不差，但對股市未必友善的報告。 消費能力仍有支撐，企業獲利短期沒有衰退式崩落，但Fed沒有急著降息的理由。加上PPI與油價走高，甚至增加升息可能性

U.S. August final-demand PPI rose 0.4% month over month and 5.4% year over year. The core measure — excluding food, energy, and trade services — increased 0.3% month over month and 4.7% year over year. The details were also important: • Energy prices rose 4.2% MoM • Diesel prices surged 24.1% in a single month • Freight transportation services increased 2% The message from this report is that upstream cost pressure has not disappeared. And this is no longer just an oil-price story. Higher energy costs are beginning to flow into transportation, logistics, and broader corporate operating costs. Fed Governor Christopher Waller had previously said that if August inflation came in hot, he would consider supporting another rate hike. If inflation continued moving toward the Fed’s 2% target, he would be more comfortable keeping rates unchanged. That makes the September 11 CPI report especially important. If CPI confirms the same inflationary pressure we are now seeing in PPI, the market is likely to trade the possibility of The Fed is hiking again. My Take The important question is no longer whether oil prices alone are moving higher. It is whether higher energy costs are beginning to spread through the economy：Energy → Freight → Corporate Costs → Consumer Prices If that transmission continues, it becomes much harder for the Fed to dismiss the inflation pressure as temporary. The next CPI report could therefore determine whether this remains an energy-driven inflation scare — or develops into a broader repricing of the Fed rate path. Make the Fed quote clearerAdd investor implications by sector

美國 8 月最終需求 PPI 月增 0.4%、年增 5.4%；核心口徑—扣除食品、能源與貿易服務—月增 0.3%、年增 4.7%。能源價格月增 4.2%，柴油價格單月暴增 24.1%，貨運服務價格也上漲 2%。 這份數據代表上游成本壓力還沒有消失，而且不只是原油價格，已經開始進入運輸和企業成本。 Fed理事 Waller 先前已經表示：如果 8 月通膨偏熱，他會考慮升息；如果通膨繼續朝 2%改善，才支持維持利率不變。 9 月 11 日公布的 CPI成為關鍵。若 CPI 同樣偏熱，市場會更直接交易「Fed重新升息」。

September 23 Market Close The S&P 500 fell 0.75%, the Nasdaq dropped 1.13%, the Nasdaq 100 declined 0.85%, and the Dow lost 0.68%. The Russell 2000 was hit harder, falling 1.77%. The VIX rose to 15.18, but it still remains well below panic territory. The 10-year Treasury yield jumped from 4.96% to 5.11%, while the 2-year yield climbed from 4.71% to 4.85%. Brent crude rebounded 3.86% to $103.08, moving back above $100. Front-month WTI gained 1.81% to $92.16. Energy was the only sector to finish higher, while small caps, real estate, utilities, and consumer-related stocks came under pressure. The market was effectively trading two things at the same time: Higher energy costs + Higher discount rates September U.S. Composite PMI rose from 56.0 to 58.4, the highest level in 62 months, while input-cost inflation accelerated to nearly its fastest pace in four years. Fed Governor Michael Barr also said that under his baseline outlook, further policy adjustment may be necessary. On the AI side, the demand story remains intact. Worthington surged about 16% on stronger expectations for data-center liquid-cooling storage tank demand. Adoption momentum around Meta Muse also remains strong. So far, there is still no clear evidence that hyperscalers are cutting AI CapEx or cancelling major chip orders. That means the AI investment cycle has not broken. But the required return threshold is moving higher again. With long-term yields back above 5% and energy prices rising, investors are demanding more proof that AI spending can convert into: Revenue → Free Cash Flow → ROIC Alphabet and Amazon both weakened, largely reflecting a combination of increased competitive pressure from Muse and valuation repricing. My Take The key message today is not that AI demand is collapsing. It is that the macro hurdle for AI valuations is getting higher. As long as hyperscaler CapEx remains intact, the infrastructure demand story still has support. But when oil rises, Treasury yields move above 5%, and discount rates increase, the market becomes much less forgiving. AI companies now need to prove not only that demand exists, but that the economics can justify the capital being deployed.

9月23日收盤： S&P 500跌0.75%、Nasdaq跌1.13%、Nasdaq 100跌0.85%、道瓊跌0.68%，Russell 2000重挫1.77%。VIX升至15.18，仍未進入恐慌區。 10年債殖利率由4.96%跳升至 5.11%，2年債由4.71%升至 4.85%。 Brent反彈3.86%至 $103.08，重新站上$100；WTI近月合約漲1.81%至 $92.16。只有能源板塊上漲；小型股、房地產、公用事業與消費類股受壓，顯示市場同時交易能源成本與折現率上升。 9月美國綜合PMI由56.0升至 58.4，創62個月新高，同時投入成本增速接近四年最快；Fed理事Barr也表示基本情境下可能需要進一步調整政策。 Worthington因資料中心液冷儲槽需求展望上漲約16%，Meta Muse採用熱度也仍在；尚未看到Hyperscaler削減CapEx或取消晶片訂單。AI投資的報酬門檻再次提高。Alphabet與Amazon走弱，主要反映Muse帶來的競爭與估值重定價。

NVIDIA 加速量子研究中心 (NVAQC) 將迎來其首個本地量子運算部署：一套 IonQ Superion 256 系統。 「我們認為這標誌著混合量子-經典計算令人興奮的新時代的開始，」我們的執行長 Niccolo De Masi 表示。 NVAQC 旨在解決將量子處理器擴展為新一代加速量子超級電腦所面臨的根本挑戰。 將 IonQ 的全端量子系統與 nvidia 的 人工智慧基礎設施直接連接，將有助於推動突破性進展，將量子位元轉化為能夠解決世界上最棘手問題的實用型大規模量子超級電腦。 我們的量子處理系統將透過 NVIDIA NVQLink 直接與 NVIDIA GB200 NVL72 系統連接。工作負載將由開放的 NVIDIA CUDA-Q 平台進行編排。 閱讀完整公告 → ionq.com/news/ionq-to-advanc… #IonQ #QuantumIsNow #QuantumComputing #QPU

Ondas 已收購 Insignito、Ottopia Defense 和 Caribou Labs，為其自主防禦產品組合增添了被動聲學感測、彈性連接和 GNSS 拒止導航技術，計劃將其整合到現有產品組合中。 $ONDS