Bitget App
Trade smarter
Buy cryptoMarketsTradeFuturesEarnAISquareMore
Accelerate! SpaceX plans to launch a "space data center" satellite next year, equipped with Nvidia chips

Accelerate! SpaceX plans to launch a "space data center" satellite next year, equipped with Nvidia chips

华尔街见闻华尔街见闻2026/08/25 00:11
Show original
By:华尔街见闻

SpaceX announced that it will launch the Starmind AI1 satellite equipped with Nvidia chips as early as next year, at least one year ahead of schedule. The partnership between SpaceX and Nvidia was officially deepened last month: at that time, Elon Musk announced that SpaceX will exclusively use Nvidia chips in the future. Nvidia also stated that SpaceX plans to use Nvidia’s Vera Rubin platform.

SpaceX is extending the battlefield for AI infrastructure into space. SpaceX announced plans to launch AI satellites equipped with Nvidia chips into orbit as early as next year, at least one year ahead of the timeline previously disclosed in its prospectus, marking an accelerated shift from the concept of "space data centers" toward practical reality.

According to MarketWatch on August 24, Musk stated on Monday that SpaceX aims to launch the first batch of Starmind AI1 satellites in the fourth quarter of 2027 and "significantly expand" the scale in the following year. This new timeline is notably ahead of SpaceX's previously disclosed 2028 deployment plans. Meanwhile, Musk revealed to investors earlier this month that SpaceX would obtain a "significant proportion" of Nvidia GPU supplies in 2027.

At the market level, this news has not boosted the share prices of either company. Nvidia's stock price fell by 2.9% on Monday and SpaceX's stock also slipped by 1.4%. Investors are focusing more on Nvidia's second-quarter earnings report to be released after the close on Wednesday, which is widely regarded by the market as a key event that could lead to major volatility.

Exclusive Partnership: SpaceX Shifts Fully to Nvidia Chips

The partnership between SpaceX and Nvidia was officially deepened last month. At that time, Musk announced that SpaceX would exclusively use Nvidia chips going forward, and both parties subsequently disclosed this exclusive collaboration.

According to Nvidia, SpaceX also plans to use Nvidia’s Vera Rubin platform to expand the AI infrastructure that supports its Grok chatbot and related services. Describing this "space-optimized" system co-designed with Nvidia, Musk said, "Our design is simpler, lower-cost, denser, and lighter than traditional racks."

This system will be produced in Bastrop, Texas, which will also be the site of SpaceX’s AI satellite manufacturing base. Analysts point out that the deep strategic alignment between SpaceX and Nvidia is forging an increasingly tight bond between the two companies’ futures.

Notably, Nvidia will release its second-quarter earnings after the close this Wednesday, an event the market considers one of the most important catalysts of the period. Cantor Fitzgerald analyst C.J. Muse recently told clients that now is the time to "go all in" on Nvidia stock.

An Ambitious Roadmap for Compute Power Expansion

According to the report, the business logic behind space data centers lies in bypassing regulatory barriers and community resistance, but the cost disadvantage is also significant. Research firm Wood Mackenzie estimates that a 1 GW orbital data center would cost around $170 billion, over three times the cost of a traditional data center.

However, supporters argue that as rocket launch costs continue to fall—driven by SpaceX’s Starship program—the economics of orbital data centers will gradually improve. Critics, meanwhile, note that for at least the next few years, ground-based data centers will still have a clear advantage in operating costs.

Morgan Stanley analyst Adam Jonas expects SpaceX’s AI-related capital expenditures to reach around $53 billion in 2026, climbing to $130 billion in 2027, equating to about $40 per watt of capital expenditure.

In addition, SpaceX’s compute power expansion plan is not limited to space. Musk stated that the company plans to accumulate more than 2 GW of AI computing power by the end of 2026, and push that number to "closer to 10 GW rather than 5 GW" by the end of 2027.

Currently, SpaceX has data centers in Tennessee and Mississippi, and is reportedly planning to build at least one more in Texas. On August 4, Musk stated:

"The entire industry is surging toward greater demand for computing power. We are rapidly expanding compute capacity to meet the internal needs brought by Grok expansion, as well as external demand from leading companies such as Google and Anthropic."

Beyond terrestrial infrastructure, SpaceX is also working with Tesla to advance a chip manufacturing project in Texas. The two companies plan to invest at least $16.8 billion in the initial phase of the Terafab project in Grimes County, Texas.

0
0

Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.

Understand the market, then trade.
Bitget offers one-stop trading for cryptocurrencies, stocks, and gold.
Trade now!

You may also like

Aiming at the low-latency AI inference market, Nvidia (NVDA.US) Groq 3 LPX enters full-scale production, with the first batch of systems to be deployed at NEBIUS (NBIS.US)

Nvidia announced on Monday that the Groq 3 LPX rack-level system has entered full-scale mass production, marking the official commercialization of this low-latency AI inference technology following the company's roughly $20 billion acquisition of Groq-related assets last year.

智通财经2026/08/25 01:06
Aiming at the low-latency AI inference market, Nvidia (NVDA.US) Groq 3 LPX enters full-scale production, with the first batch of systems to be deployed at NEBIUS (NBIS.US)

Goldman Sachs Head of Hedge Fund Business: Oil prices are more important than the Jackson Hole Fed meeting; Nvidia is the focus this week

Privorotsky, Head of Hedge Funds at Goldman Sachs, issued a warning this week: the movement of oil prices may have a greater impact on the stock market than the Jackson Hole Symposium—lowering energy prices might be the fastest way to stabilize the bond market. Although Nvidia’s earnings are expected to be “absurdly strong,” there are undercurrents of concern regarding an AI financing bubble, rising memory costs, and valuation compression. He suggests going long on financials and industrials, holding “S&P + gold,” waiting for the right opportunity to short bonds, and for semiconductors, waiting until earnings catch up with valuations before seeking a breakout.

华尔街见闻2026/08/25 00:56

The U.S. pressures Korean semiconductors to build factories in the U.S., expresses dissatisfaction with South Korea's “Honam Semiconductor Cluster Plan”

According to reports, officials from South Korea's ruling party revealed that the US side has explicitly requested Korean companies to build memory factories in the US to ensure supply. After the announcement of a mega semiconductor project in Hunan, pressure on semiconductor investments has increased significantly, with accusations that Korean companies have delayed implementing their semiconductor investment plans in the US. The South Korean government plans to announce the first batch of US investment projects at some time in September.

华尔街见闻2026/08/25 00:56