
#SpaceX99%ValueFromAI
About SpaceX99%ValueFromAI
Musk told SpaceX staff that AI revenue should surpass all other businesses in September. He targets 10 GW of compute by end-2027, which he estimates could generate $300B-$500B annually. His "train on Earth, infer in space" plan would combine Starship capacity, Starlink and AI compute into one infrastructure stack. Musk expects AI to drive 99% of SpaceX's value in five years. These remain management forecasts. Can AI revenue support its valuation, and are capex and execution risks fully priced?
Populares
Más recientes
SpaceX99%ValueFromAI Publicaciones populares
Musk is turning SpaceX into an AI infrastructure company.
He told staff that AI revenue could surpass all other SpaceX revenue combined in September. He also expects AI to account for 99% of the company's value within five years.
SpaceX reported $2.56B in Q2 AI revenue, up 247% YoY. But Musk's target is far larger.
He says compute could expand from roughly 1.4 GW today to 10 GW by end-2027, more than sevenfold. Using Musk's $30 to $50 per-watt revenue estimate, that would imply $300B to $500B a year.
Cloud contracts show demand is not theoretical. Anthropic and Google have agreed to buy access to SpaceX's terrestrial compute capacity. The Google agreement covers roughly 110,000 Nvidia GPUs and $920M in monthly payments from October 2026 through June 2029.
But those revenues depend on delivery. If SpaceX misses its Sep 30 GPU commitment, Google can terminate the agreement or accept fewer GPUs at proportionally lower fees after a one-month grace period. After Dec 31, either party can terminate with 90 days' notice.
Chip supply is another variable. Musk says SpaceX will build its AI infrastructure exclusively on Nvidia chips, citing Vera Rubin as its preferred architecture. That could simplify the technology stack, but it also increases reliance on one supplier.
The broader strategy connects several layers:
· Earth-based clusters for training
· Starlink for global connectivity
· Starship for orbital deployment
· Space-based compute for future inference
The stack is not equally mature. Today's AI infrastructure revenue is still tied to terrestrial compute, while commercial-scale orbital inference remains in early development and has not been proven.
So this is no longer only a rocket or satellite-internet story. It is a test of whether SpaceX can turn compute, connectivity and launch capacity into one scalable AI network.
Q2 revenue provides an operating base, but the 10 GW target, $300B to $500B estimate and 99% value claim remain management forecasts. What will decide this story first: customer demand, chip supply or execution?
#SpaceX99%ValueFromAI
SpaceX is going to start looking crazy cheap if you believe what management is saying.
$SPCX today is still ~$1.9T market cap / ~$1.8T enterprise value.
On the last twelve months of sales (~$23B), that’s roughly ~80x for EV/Sales.
On today’s numbers… it is expensive, ofc.
But these multiples will start compressive drastically if what Elon said on the earnings call becomes true.
1/ $100B ARR by December 2026
“It’s not a question mark… that’s what we would achieve if we basically did nothing.” - Elon Musk
Hold today’s EV fixed and put $100B in the denominator:
The multiple of ~80x becomes → ~18x EV#CPIEasesHikeBets #AIInfraEarningsWatch #SpaceX99%ValueFromAI

It’s actually happening: Video footage of construction for SpaceX’s project to put AI Data Centers in orbit
These Data Centers in space won’t take up earths resources like electricity and water because they will use solar power and be cooled by space
“The SpaceX campus in Bastrop County is pushing for an ambitious project that would build what it's calling the AI1 satellite, a data center in space. There are visible signs construction is underway after the tech company announced this large expansion project where the Starlink facility currently sits it's being called the GigaSat factory
— The company plans to launch these satellites into Earth's orbit with SpaceX Starships. The company says this tech will be powered by solar energy, and the cold temperatures in space will keep it from overheating”
The goal is large scale meaningful production by the end of 2027, supporting an annualized rate of ~1 GW of orbital AI compute. This means thousands of satellites per year, with even more to scale much higher afterward
Musk is turning SpaceX into an AI infrastructure company.
He told staff that AI revenue could surpass all other SpaceX revenue combined in September. He also expects AI to account for 99% of the company's value within five years.
SpaceX reported $2.56B in Q2 AI revenue, up 247% YoY. But Musk's target is far larger.
He says compute could expand from roughly 1.4 GW today to 10 GW by end-2027, more than sevenfold. Using Musk's $30 to $50 per-watt revenue estimate, that would imply $300B to $500B a year.
Cloud contracts show demand is not theoretical. Anthropic and Google have agreed to buy access to SpaceX's #CPIEasesHikeBets #AIInfraEarningsWatch #SpaceX99%ValueFromAI
#马斯克称AI将占SpaceX价值99%
马斯克又“画饼”?99%价值靠AI,SpaceX到底是个啥公司?!😳
刚刷到个新闻,真的给老马跪了!跪键盘的那种。😂
他在SpaceX全员会上放话,说五年后AI要占公司99%的价值!我没看错吧?一家造火箭的、铺星链的公司,最后的核心价值竟然全在AI上?
老马还列了个详细的数据:9月AI收入超其他业务,明年底要弄10吉瓦算力,算下来对应3000亿-5000亿美元年收入!这数字,比现在很多国家的GDP都高吧?
他还搞了个“地面训练、太空推理”的部署套路。听起来很高大上,但网友都在评论区炸了,说他又在“画饼”。
我是这么想的:如果老马真的把星舰运力、星链网络和AI算力整合成功了,那这就是个前无古人的超级AI基建巨头。但这个“如果”也太大了!现在的估值里,到底有多少是在为这个巨大的“饼”买单?算力扩张要花多少钱?执行风险有多大?这些都被充分定价了吗?
我就想问问兄弟们:这波你信不信?如果信,对应的代币 XSPCXSPCX 是不是该直接冲?如果不信,老马这次会不会直接“贾跃亭化”?
$XSPCX $SPCX $BTC

#马斯克称AI将占SpaceX价值99%
99%是AI,1%是火箭,我的空单在中间
$SPCX 站上146了,从108算起涨了35%。我的空单浮亏300U,-1925%,没走。
马斯克说话了。全员会上说,AI收入预计9月超过其他业务总和。明年底10吉瓦算力,按他的测算对应3000亿到5000亿年收入。五年后AI占SpaceX价值的99%。
99%是AI,1%是火箭。
5000亿年收入什么概念?英伟达去年营收600亿。一个还没商业化的AI业务,五年要做到5000亿。不是增长,是换物种。
但市场信了。从108到149,40%涨幅,马斯克一句话喂出来的。一个多月前228的SPCX讲的是火箭和星链,现在146的SPCX讲的是AI和5000亿。同一个标的,换了个故事,价格就能回来。
换故事不需要财报,一场会就够了。
我不评价这个判断对不对。但我确定一件事——故事能拉盘,也能砸盘。故事不需要兑现,市场信就行。但故事撑起来的价格,需要数字来验证。
空单还在,不是不信AI,是不信5000亿会从一场全员会上长出来。我在等财报,等订单,等数字说话。
他说99%是AI,我还在那1%里等。故事能拉盘,但不能当饭吃。
Instantánea del 13 ago 2026, 13:28
🚀火星太难上,马斯克开始画AI大饼
SpaceX卖火箭的,突然说自己99%价值在AI?
马斯克在全员会上放话—明年底搞到10吉瓦算力,对应3000-5000亿美元年收入,5年后AI贡献公司99%价值。
逻辑上有个明显矛盾—SpaceX的核心是火箭、星链、火星殖民,怎么突然变成AI公司了?说白了——火星计划太烧钱,短期看不到商业兑现,只能拿AI讲故事撑估值。
财报里183亿美元的季度资本开支已经让市场吓一跳,现在又加码AI算力。如果SpaceX是火箭公司,投资人会盯着发射收入和星链用户,如果它把自己包装成AI公司,故事就变成了“算力租赁巨头”——这个赛道目前市场给估值确实更慷慨。
火星上不去,AI饼先画上,别管能不能实现,故事得先讲下去。 🚀
#马斯克称AI将占SpaceX价值99%




