Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko) header
Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko) profile picture

#交易紀錄與日常發現 Dream investments of buffett type #ASTS #ABTC “I hope my followers don't mind me enjoying a little NSFW content from time to time”

Taipei City, Taiwan
Joined May 2017
Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)Photo by Mojo YoYo 🅰️🛰️📱🅰️ (@Lo_Do_Ko)
僅遵守此份調查報告維持原持有之股份並在每次回調時觀察狀況適時買入🐦
$ASTS The big boss has revised down the number of satellites needed to break even (previously 20 instead of 25). This has some profound implications on when the company will get to cashflow breakeven. Basically all it now has to do: > Build 16 satellites > Launch them via New Glenn (2x) or Falcon 9 (4x), probably a combo of 1x New Glenn (8 birds per flight) launch and 2x Falcon 9 (4 birds per flight) launches before end of 2025 > MNO's will handle the go live and deployment model to their hundreds of millions of subscribers How realistic is this before end of 2025? Based the Q3 2024 earnings call I think this is quite realistic, although at Q4 2024 earnings call we will probably get the fine details, however by that point the stock price might already go on a run, since then institutions will do the napkin math I do below as well. Below is a long assessment, but the TDLR is that this company is about to hit the sweet spot of the S-curve and the market is not yet fully understanding how profound these implications will be. The road to cashflow break-even requires: Launch First in March/April via ISRO the first block 2 bluebird will go up as a sort of "test" satellite for the new Block 2 generation. Judging that last time they require 2-3 months of calibration it is likely to assume the subsequent two launches with SpaceX are likely set in Q3 2025. Considering ISRO's and SpaceX track record with the vehicles I consider these "slam dunks". This would mean by end of Q3 then ASTS has 14 birds (plus Bluewalker) in orbit and operational or a close equivalent. Abel then confirmed one New Glenn launch is booked for 2025 as well, meaning the constellation would get to 22 birds, which would then mean cashflow break-even. With Vodafone's recent announcements of end of 2025 / beginning of 2026 I think an European launch for intermittent service will get the company to breakeven. Since New Glenn is a new vehicle with now just one (succesful) launch there is definitely some uncertainty in that launch that is bookmarked for end of 2025, but management confirmed they can backfill this with SpaceX. I think it is likely they have an option in the contract. Might cause a quarter delay. If I look at it from launch perspective end of 2025 is the everything goes as planned scenario and Q1 2026 is the reasonable prediction factoring in some delay type of scenario. That would mean the company is just one year away from cashflow breakeven. Satellites The company confirmed in the call that they have scaled the production to meet the 45-60 birds target in 2025 and 2026. Pointing out the building blocks of block 1 (Microns) are for the most part identifical to block 2. Lessons have been learned on block 1 and those will be put to fruition on block 2. In earlier calls we also got the number of production of 2 birds per month that has the goal of scaling to 6 per month in 2026. Basically if we do the conservative math AST building 2 birds per month for nine months that would get them to have 18 birds ready in Q3 2025, which is in time for the launch campaign. This seems doable, especially after the convert were management said they are going to fast track satellite production even further then iterated in their Q3 2024 earnings. Will need to get more details on this in the next call. ASIC production seems also on schedule, but again birds can also be launched with the FPGA configuration if there are any delays. Based on my assessment I conclude that from satellite production perspective nothing will be in the way for a end of 2025 launch. Q1 earnings call will likely be a confirming event with more details MNO Judging on recent MNO action it is obvious they will be ready when AST is ready to launch the service, that's a no brainer for me, so no delay from this angle. Cashrun way After the convert the company has a close to 1 billion warchest. Based on management comments we can calculate the capital expenditures and conservatively assume nothing is paid yet: (which is not entirely true) > 16 birds costs on average 20 million to build (source: Abel Avellan) = 320 million. > New Glenn is rumored to cost $100 million (AST might get some discounts due to signing before first launch), Falcon 9 costs 70 million. ISRO probably in the 20 million ballpark = 260 million. > Opex burn rate is 30-35 million, lets asumme 40 million to be conservative = 160 million. Conclusion Basically to get more then 20 birds up this would cost the company 580 million, with 320 million capex and 160 million of regular OPEX. With the close to 1 billion war chest this goal is easily obtainable without any additional funding or prepayments (which is extremely unlikely) and even if you factor in 2 quarters of delay there still will be sufficient cash. To conclude this basically means AST has a very credible path to cashflow break-even end of 2025 or Q1 2026. I think the stock could easily more then double if this goal is attained, which would mean from today's price within a year you get a double. But Tim management is going to accelerate again right? Yes, I think what happens is that management before convert was very focused on that 20-25 bird scenario, but I think now management is really going to put the pedal to the medal to get as fast as possible to continous coverage of 40-60 birds. Meaning instead of optimizing the cash position everything will be focused on accelerating sattelite production and probably buying additional launch capacity from SpaceX. We will know more during the Q4 call, but this is how I think it will play out. Very excited about the next 1-2 years. It will be a game changer for the company and my portfolio :) Attachments Relevant extracts from the Q3 2024 earnings call: Launch "As Abel and Scott detailed, during the early fourth quarter, we secured launch contracts with providers to enable us to launch up to approximately 45 Block 2 BlueBird satellites with options for additional launch vehicles up to approximately 60 Block 2 BlueBird satellites through 2025 and 2026" Launch schedule "So we will -- the cadence will be probably 1, 4, 4, 8, 8, 8 until we get to 60. And the New Glenn is designed to be reused 25 times and the multitude of them being built and with the first one scheduled to actually be this year." Operating costs burn rate "Consistent with the first, second and third quarters of 2024, we estimate that our adjusted cash operating expenses for the fourth quarter, excluding some remaining ASIC costs, will come in within a range of $30 million to $35 million as we continue to scale production of our Block 2 BlueBird satellites in preparation for our 2025 and 2026 launch schedule." ASIC chip integration for Block 2 birds (Block 2 will both have FPGA and ASIC chips) "And post that later in '25, mid-2025 to third quarter 2025, we will be including our ASIC in our subsequent launches." Scaling up Block 2 production "In Midland, we basically have scaled up to have the ability to produce the Microns that are required for Block 2 and then basically meet our target of 60 satellites during '25 and '26. We have done the vast majority of all the investment required for that. And basically, the same technology that we did launch in Block 1, it will be what it will be used on Block 2 satellites as it relate to the Microns for subsequent Block 2 satellites. So they're basically the same Microns, the same infrastructure for the next few launches. using our current Micron that is successfully operating now in Block 1." & "We currently expect our average costs of direct materials and launch expense per satellite for our Block 2 constellation to be in the range of $19 million to $21 million, an increase from our prior estimate of $16 million to $18 million per satellite as a result of actual launch costs recently contracted"
4,203
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
Let's be serious about $ASTS Look at this chart. Moonshot ascension followed by brutal drop of 83%... Something about expecting consumers to understand an evolutionary technology... Oops - that's the chart for $AMZN from back in the day
Soyons sérieux quant à Bitcoin. Regardez ce graphique. Une hausse violente de 65% puis une chute brutale de près de 40% en quelques mois. Avec une telle volatilité, Bitcoin est à des années lumières de devenir une réserve de valeur mondialement reconnue, et encore moins un actif de réserve mondial. Comment voulez-vous que l’on soit pris au sérieux si l’on continue de prophétiser de pareilles absurdités ? Les investisseurs professionels, les entreprises cotées, et même les familles et PME, ne se risqueront jamais à détenir de l'épargne de long-terme dans un actif aussi volatile. Vous réalisez à quel point l’on passe pour des idiots fanatiques et incultes lorsque l’on ose avancer, qu’un jour, les institutions de crédits et les banques centrales n'auront d’autre choix que d'utiliser cet objet spéculatif comme actif de réserve ? Oui cet actif au porteur est rare, neutre, vérifiable, et exempt de risque de contrepartie. C’est indéniable. Mais comment ignorer son immaturité rédhibitoire qui, seule, peut expliquer une telle volatilité ?? Il s’agirait d'arrêter de se ridiculis.. Oups... Pardon, oubliez ce que vous venez de lire. Je me suis trompé. C’est le graphique de l’or, pas celui de Bitcoin.
13
25
3
304
18,524
看來稅是下一隻腳
Wall Street has been peddling abusive tax schemes to the wealthy. The IRS will come down hard on both the firms promoting them and the investors using them. Firms should win business by delivering better returns, not by dreaming up ever more outrageous tax gimmicks and exposing their clients to the consequences. Hardworking Americans put in an honest day’s work and pay what they owe. We want capital growing businesses and creating jobs, not funding the next tax dodge.
14
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
2022年台積電盈利無懼升息暴增 70%,股價卻照樣一度腰斬,PE 從 26 倍殺到 10 倍,結果 2023年盈利真的開始下滑,市場提前一年就知道。 如果民主黨真的暫停尚未取得許可的 AI 資料中心,2027 年項目還能繼續,2028 年新增建設卻會直接停擺。 難道你以為投資者會乖乖等到 2028 年才逃命?
Downcycle 別看 PE,要看 PB
12
10
1
179
19,127
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
我最新的產業調查顯示,Google 下修 2027 年、由 Broadcom 設計的 TPU 8i 系列共約 100 萬顆(以單價較高的 8i Max / Whalefish 為主)。此下修約佔 2027 年 TPU 8i 系列出貨量約 20%。 本次調查也指出,Google 預計增加 2027 年由聯發科設計的 TPU 8t 系列出貨,但細節尚未確定。Google 近幾個月反覆強調前沿模型研發優先取得算力,此次 8t 與 8i 系列出貨比重的改變,正是在供應鏈緊張與資料中心供電受限下所做的調整。
8
35
2
522
68,226
快點結束吧
JUST IN: President Trump just put Iran on notice from the White House podium. "They want to make a deal, but they don't want to make a deal that's acceptable to me," Trump said, adding, "We want to make a real deal, not a fake deal." Then he dropped the hammer: "They've been bulls***ing people for 51 years." No more games. 🔥🇺🇸
20
削減開始
My latest industry survey indicates Google has cut 2027 shipments of the Broadcom-designed TPU 8i series by ~1M units (mostly the higher-priced 8i Max / Whalefish). The cut amounts to roughly 20% of 2027 TPU 8i series shipments. The survey also indicates Google plans to increase 2027 shipments of the MediaTek-designed TPU 8t series, though details aren't final yet. Google has repeatedly stressed in recent months that frontier model development gets first call on compute, and this shift in the 8t and 8i shipment mix is precisely the adjustment it's making amid a tight supply chain and data center power constraints.
1
82
之後要等美國說明到底怎樣才算是真正的走私,不然你每一次看到中國企業有美國禁止的產品,每一次就是對公司的不信任,你只是在扼殺美國本土的企業價值 走私這個問題,不是做空機構的玩具 Fuckyou走私和禁令的空頭
🚨 HERE WE GO $AMD: EXPOSING HUNTERBROOK SHORT REPORT ON AMD HERE'S WHAT THEY'RE NOT TELLING YOU. I went through their report and I found several places where their conclusions go further than the evidence. Not to forget guys, Hunterbrook Capital disclosed that it was SHORT $AMD when the report was published. 1. $AMD firmware found at a blacklisted Chinese company. Hunterbrook found six newer AMD firmware versions at IEIT after its March 2025 blacklisting. Hunterbrook itself admits it doesn't know how IEIT obtained that firmware or whether a valid license existed. Finding AMD code doesn't mean AMD illegally supplied it. 2. The 99.94% firmware similarity. Hunterbrook found nearly identical firmware at IEIT, Aivres, Kaytus and Maginfra very interesting finding but the servers appear to share the same underlying design. Similar firmware is hardly surprising in that situation this alone doesn't establish illegal chip transfers. 3. The $5.6 BILLION shipment figure. Hunterbrook highlights $5.6 billion in technology shipments from Aivres to Southeast Asia but more than $3 billion involved NVIDIA-based systems so how much was actually AMD-related? The report doesn't establish that. You can't take an intermediary's total shipments and treat them as AMD's exposure. 4. AMD's $67 MILLION licensing income. Hunterbrook connects this to AMD's historical Hygon ventures in China but these agreements existed before the companies were blacklisted in 2019 even Hunterbrook acknowledges the payments don't establish an export control violation. Receiving royalties isn't the same as transferring restricted technology. 5. $AMD's 22.4% China revenue versus $NVDA's 9.1%. This comparison bothered me. AMD's figure reflects revenue billed to China and Hong Kong. NVIDIA uses customer headquarters. Different reporting methods, different fiscal periods. AMD's number includes CPUs and other products, not just restricted Instinct accelerators. The 22.4% figure doesn't tell us how much revenue is actually at risk. 6. The $4.28 BILLION Singapore exposure. Hunterbrook points to AMD's Singapore revenue as new export restrictions take effect fair concern. But billing a customer in Singapore doesn't establish where the chips ultimately go or whether every product requires an export license where's the breakdown of affected products? 7. AMD's relationship with Tongfu. AMD purchased around $2 billion in assembly, testing and packaging services from Tongfu-related ventures in 2025 it's a real supply chain dependency and I wouldn't dismiss it but it doesn't establish that AMD diverted chips to restricted Chinese companies. 8. The Cancon server listing. Hunterbrook found a Cancon server in AMD's February 2026 catalog. Cancon is listed as an alias of blacklisted Suma Technology a catalog listing isn't proof that AMD shipped restricted chips without authorization. Hunterbrook has uncovered relationships and compliance questions that deserve answers. I wouldn't dismiss those findings but there's a big difference between identifying potential export-control risks and proving that AMD knowingly violated those rules. The report hasn't established the latter. I'd like $AMD to address the firmware and intermediary questions directly. Until then, we shouldn't treat every suspicious connection as evidence of wrongdoing esp when the firm publishing the allegations has disclosed a short position in the stock.
26
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
Anduril 創辦人拉奇(@PalmerLuckey)說,黃仁勳下台,是中國贏了。11 月 18 日,他在《紐約時報》DealBook 峰會上被主持人 Andrew Ross Sorkin 問到這件事,他的說法是:中國想要黃仁勳走,所以他要黃仁勳回來。 拉奇在晶片出口上跟黃仁勳立場相反,他認為美國不該把晶片賣給中國,但也說黃仁勳的理由說得通。他接著回應一種反駁,有人說在幫中國的明明是輝達。拉奇說,想買輝達 GPU 的是中國公司,中國政府不想讓他們買,因為北京要建立自己的晶片產業,不想依賴美國;他說現在有 60 萬顆 H200 擱在港口,中國不給進口許可。路透 11 月 5 日報導過這個數字,兩位知情人士說這批 H200 存放在香港,等北京批准大量進口;另外,幾家中國網路大廠近幾個月已經各拿到大約一萬顆。 所以拉奇說,黃仁勳站在中國 AI 公司那邊,不是中國政府那邊,中國政府很高興他走了。他也說黃仁勳是從草創時期一路打造輝達的人,大家願意跟輝達做生意,靠的是他建立的人脈,在這個關鍵時刻沒有人能取代他。他同時補了一句,黃仁勳不該對妻子不忠,這件事黃仁勳在辭職前寫信向員工承認過。 輝達 11 月 12 日宣布黃仁勳辭職,11 月 17 日的財報法說會上,代理執行長兼財務長 Colette Kress 說,輝達已經停止把需要出口許可的 Hopper 資料中心產品運到中國。H200 屬於這一類,輝達沒有說停運跟黃仁勳辭職有關。董事長 Harvey C. Jones 在同一場法說會上表示,董事會不打算讓黃仁勳回任。資訊截至 2026 年 11 月 19 日。
8
44
3
278
22,661
審查完老黃,接著審查蘇媽
BREAKING: AMD stock, $AMD, crashes -13% in 30 minutes after Hunterbrook Media accuses the company of illegally selling chips to China. Hunterbrook Capital also disclosed a short position in AMD.
26
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
Nvidia召開內部員工大會,新CEO再度重申不會讓黃仁勳回歸,要加薪留才穩定軍心 Kress 開場直接說「Let’s move forward」(讓我們繼續向前走),強調公司會安然度過,並重複財報會上的「We’ve got this」。 她明確表示最重要的是照顧每一位員工,讓大家對公司有信心、不用擔心薪酬。 #薪酬與股權(員工最關心的議題) 事前收到約 700 個員工問題,薪酬是第一熱門主題。Kress 宣布將年度考核/調薪與額外股權發放從原本的 3 月提前到 12 月,遠早於常規時程。 她說:「我們最想做的,就是照顧每一位在這裡的員工,讓你們對這家公司感覺良好,不用擔心你們的薪酬(comp)。」這被視為穩定人才、防止流失的關鍵措施。 #董事會透明度與 CEO 接班 第二多的問題是希望董事會更透明。領導層重申董事會已一致接受黃仁勳辭職,無意讓他復職,正在尋找永久 CEO(內外部人選都考慮)。 這與部分員工內部外連署、請願希望黃仁勳回歸的聲音形成對比;黃仁勳本人也曾表示若董事會認為符合股東、員工、公司與美國利益,他願意回來。 #業務延續與未來方向 主管們強調路線圖、客戶關係與執行力都沒有改變: • 客戶與夥伴在領導層變動後仍表達支持。 • 需要把過去依賴黃仁勳個人關係的「executive touch」擴展到更廣泛的高階團隊。 • 加速從晶片公司轉型為完整「AI factories」供應商:從晶片 → 系統/機架 → 明年起的容器化 pods(一次交付 8 個機架)→ 最終客戶可直接訂購完整資料中心。 • 供應商關係穩固,建立在互利基礎上。 • 持續支持開放與封閉模型,將開放模型視為下一代開發者平台(Nemotron、Groot 機器人、Cosmos 模擬、Alpamayo 自動駕駛等),並表示最終會收購 Hugging Face。 • 遊戲業務仍重要(每年出貨數千萬張顯示卡,服務約 2.5 億使用者)。 • 維持「Top 5 things」內部溝通機制,讓問題能跨層級被看見。
41
7
1
256
51,833
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
$ASTS: As always, a terrific post from Mike. If you are going to leave with one take, it should be that nearly all the bear scenarios are already priced in. From Mike: “For $10.1B of enterprise value to be the right answer, you have to accept the following simultaneously: -Starlink Mobile dominates D2D globally and AST becomes an also-ran despite 60+ MNO agreements covering 3B+ subs; -US government radar and comms stalls below $500M annually despite the awarded run-rate and the FCC authorization and the first radar satellite going into space in Q1; - Japan J-LEO doesn't finalize and the Q4 2026 Rakuten Saikyo launch slips materially despite the eighteen-month public commitment; - The European four-MNO LOI never converts and the Brazilian coalition never finalizes; - The spectrum portfolio has no standalone value despite the SpaceX / Grain price discovery; - BlackLift fails to deliver or slips multiple years past late 2028” That’s a lot of negative priced into the stock.
7
5
151
17,025
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
$ASTS - Market is sleeping on AST pulling forward their integration of the ASIC chip to BB34 versus prior plans of BB51.... Governments contracts are ramping up quicker than the market anticipates! 👀 - Deutsche Bank
$ASTS - The market is valuing AST SpaceMobile at $11B yet gov't around the world are literally throwing money at them.... - Deutsche Bank
6
9
1
114
5,327
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
Over the past week, I’ve been overwhelmed by the love and support of friends, industry colleagues, and people I've never met. Thank you to everyone who has shown kindness to my family and me. It means the world to me. I came to the United States 53 years ago, and becoming an American citizen is one of the greatest gifts and proudest moments of my life. NVIDIA has been my life’s work. Building it into the company it is today alongside my colleagues has been a privilege and an honor. Nowhere but in America could a company like NVIDIA be possible. I love our country, and I love NVIDIA. I did not live up to the standards expected of everyone at NVIDIA, and I have made serious mistakes in my personal life. I created problems for the Board that I deeply regret. I own my mistakes completely, and I am committed to cooperating with the Board’s independent investigation. Since my departure, I’ve seen an outpouring of calls from NVIDIA’s employees, investors, business partners, and customers for me to return as CEO. Many continue to reach out to me directly as the company navigates complex technology, market, competitive, and geopolitical dynamics. NVIDIA is a once-in-a-generation company. If the Board of Directors decides that my return is in the best interest of our shareholders, our employees, the company, and America, I would be honored to answer these calls and return as CEO. Thank you again to everyone who has shared their good wishes. I am profoundly humbled by your support. To my family: I love you, and I will spend the rest of my life being worthy of yours. Jensen
61,370
121,791
24,523
1,175,269
59,430,103
Lutnick out Bessent & Rubio In
Talk about timing Commerce Secretary Lutnick’s family firm “Cantor Fitzgerald” just disclosed latest Q3 holdings The firm cut its $NVDA position 28%, avoiding a 32% drop in the stock During same quarter the firm was selling, Lutnick was slowing attempts that prevented chip smuggling to China Cantor called the NVIDIA trade “client swap transactions” And Lutnick’s sons run the firm since the Commerce Secretary took office
24
水啦
$ASTS: 🚨 AST NOTIFIES FCC THAT IT PLANS TO BEGIN INTERNATIONAL SERVICE IMMINENTLY On November 13, AST SpaceMobile formally notified the Commission that it will begin international direct-to-cell operations “imminently” with its MNO partners. More importantly, AST certifies that ALL necessary national regulatory authorizations have ALREADY been secured in the countries identified in its filing. Cc: @CaptainScuzzy
1
95
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
美國商務部BIS工業安全局在今天凌晨發佈了長達132頁的最新半導體禁令,未來AI晶片的出口禁令將擴大到新加坡、馬來西亞、泰國、印尼、巴西五個國家。 (本來還有南非跟挪威) 未來這五個國家必須取得BIS的許可,才能進口受管制的AI晶片,使用這五個國家算力的美國終端客戶除外。 未來全球的AI資料中心只要輸出1200萬以上TPP算力,到中國(含香港)、澳門,以及受美國武器禁運管制的D5國家,一律需要申請許可。 1200萬TPP相當於750顆輝達H100晶片同時運算,這是小型入門級資料中心的水準,也就是說,BIS禁令之下,中國公司僅能利用遠端運算進行小規模推論,但是無法用這種算力水準訓練AI模型。 跟台灣有關的部分:晶圓製造廠和封裝廠,未來在為來自這五個國家的IC設計公司出貨時,都需要向BIS申請許可證。 (Rubio大獲全勝啊)
34
92
6
747
32,199
欸不是,我們討論了黃樂一個禮拜而且她還讓Jensen 失業的同時失去所有的名聲,結果你跟我說她去年2025 /8月就已經離職,開我們玩笑嗎?那之後的婚外情關Nvidia 屁事
Singapore-based tech company Megaspeed said that Alice Huang, a former executive and alleged Chinese spy whose relationship with Nvidia CEO Jensen Huang led to his ouster, left the company in August 2025. Read more: wapo.st/4t4K5Tk
1
60
Mojo YoYo 🅰️🛰️📱🅰️ retweeted
SpaceX 與 Singulus 的太空運算合作,可望推升太陽能光電 - SpaceX 與 Singulus 合作案(11月10日):Singulus Technologies(SNG.DE)將為 SpaceX 的 Starmind 太空 AI 運算衛星,供應統包式 HJT(異質接面)太陽能電池製造設備。首批訂單超過6,000萬歐元,第一條產線在2027年上半年交機,未來有機會擴大到年產能超過100GW。這項合作延續2026年3月的夥伴關係,SpaceX 也提供了資金支持。 - 太空太陽能結合 AI 運算,是太陽能光電的新成長動能。AI 資料中心用電需求升高,太陽能建置速度快(6至15個月),比天然氣和核能更有優勢。美國能源資訊署(EIA)預估,2026年美國公用事業級太陽能新增43.4GW,年增60%,約占全美計畫新增裝置容量的51%。美國能源部(DOE)也在支持太空太陽能。 - HJT 是太空太陽能的首選技術:在太空中的轉換效率有20%至24%,壽命終期仍能保有約95%的發電功率,成本(每瓦0.4至0.6美元)低於砷化鎵(GaAs),還能沿用現有矽晶基礎設施和超薄晶圓。在抗太空輻射、重量和量產擴充性上,都優於 PERC 與 TOPCon。 供應鏈受惠者
1)設備:Singulus Technologies(SNG.DE)
2)模組:元晶(TSEC, 6443.TW)
3)材料:倉和(Brave C&H Supply, 6538.TW),提供先進金屬化網版,可印出更細的線路並減少銀用量
SpaceX-Singulus Space Compute to Boost Solar PV •SpaceX–Singulus deal (Nov 10): Singulus (SNG GR) to supply turnkey HJT solar cell manufacturing equipment for SpaceX’s Starmind space-based AI computing satellites. Initial order >€60mn; first line delivery 1H27; potential expansion to >100GW annual capacity. Builds on March 2026 partnership with SpaceX financial support. •Space solar + AI compute is a new growth driver for solar PV. Rising AI data center power demand favors solar due to fast build times (6–15 months) vs. gas/nuclear. EIA sees US utility solar +43.4GW in 2026 (+60% YoY, ~51% of planned capacity). DOE also backing space solar. •HJT is the preferred tech for space solar: 20–24% space conversion efficiency, ~95% end-of-life power retention, lower cost ($0.4–0.6/W) than GaAs, and leverages existing silicon infrastructure/ultra-thin wafers. Better than PERC/TOPCon for space radiation, weight, and scalability. Supply chain beneficiaries 1Equipment: Singulus (SNG GR) 2Modules: TSEC (6443 TT) 3Materials: Brave C&H Supply (6538 TT) — advanced metallization screens for finer printing & lower silver use #space #solar #PV #solarcell
1
2
23
5,236