SpaceX AI Revenue Jumps 9-Fold, Supermicro Raises Sales Forecast by 27%
Lam Research Equipment Sales Rise 35%; Meta Faces Mounting Investment Pressures
"While this does not resolve every question surrounding AI return on investment, it is a good start."
Angelo Kourkafas, Senior Global Strategist at Edward Jones, told Reuters on the 21st (local time), offering this assessment of the recent earnings reports from US artificial intelligence-related companies.
As global stock markets tremble amid debates over AI overinvestment, solid revenue growth across AI services, advertising, servers, and semiconductor equipment was confirmed in the earnings of major US corporations. SpaceX's AI business revenue skyrocketed more than ninefold in a single year, while Supermicro raised its quarterly sales forecast by 27%. Lam Research also saw a 35% increase in new semiconductor equipment sales. Meanwhile, Meta maintained growth in its advertising business, but its free cash flow turned negative due to massive capital expenditures.
On the 21st on Wall Street, a rebound materialized, led heavily by AI-related equities. SpaceX rose 7.8%, Supermicro gained 9.6%, and Dell advanced 5.3%. Buying sentiment flowed back into technology stocks that had previously weakened over concerns regarding AI investment profitability.
SpaceX AI Revenue Up 9-Fold... Meta Faces Escalating Investment Pressures
SpaceX announced on the 20th that its third-quarter revenue reached $14.6 billion, up 161% from the same period last year, with net income coming in at $4.4 billion. Revenue from its Superintelligence (SI) division, which oversees the AI business, surged approximately 9.3 times to $7.56 billion, accounting for more than half of total sales. The AI division's operating profit also turned a profit at $431 million.
The strong performance was driven by xAI—acquired last February—along with expanding cloud service contracts and growth in the AI software business. AI compute infrastructure expanded to 1.7 gigawatts (GW) as of the end of September. Third-quarter AI capital expenditures totaled $16.444 billion, more than double the division's business revenue over the same period.
Meta confirmed the effectiveness of leveraging AI within its legacy advertising business. Third-quarter revenue rose 26% year-on-year to $64.6 billion. Ad impressions increased by 10%, while the average price per ad climbed 14%. AI-driven recommendations and improved advertising efficiency underpinned this top-line growth.
Earnings per share (EPS) stood at $3.83, missing market expectations, but reached $7.40 when excluding approximately $9.89 billion in litigation-related expenses, beating consensus estimates.
Meta raised its full-year capital expenditure outlook to between $135 billion and $145 billion. Third-quarter free cash flow registered a negative $2.12 billion. Bank of America (BofA) projected that massive expansions in AI investments could weigh on Meta’s per-share earnings next year.
Financing burdens at cloud provider CoreWeave also heightened market anxiety. Yields on certain unsecured bonds of CoreWeave, which carries roughly $50 billion in debt, climbed to the 15% range. Aside from revenue growth in the AI business, scrutiny over heavy capital outlays and debt burdens is intensifying.
Server Sales Outlook Raised by 27%... Surging Demand for Semiconductor Equipment
US AI server manufacturer Supermicro released its preliminary sales outlook for the July–September quarter on the 20th, raising it 27% from previous estimates to between $18.5 billion and $19.5 billion. This significantly exceeded the market consensus of $15.23 billion and represents roughly 3.8 times the revenue of the same period last year. The gross margin forecast was also upgraded from 10.4–10.8% to 12.0–12.5%, simultaneously elevating expectations for both top-line growth and improved profitability.
Semiconductor equipment maker Lam Research posted quarterly revenue of $8.219 billion, a 54% increase year-on-year. System revenue, centered on new equipment, surged 35% quarter-on-quarter to $5.733 billion. Chief Executive Officer Tim Archer cited rising AI demand and increasing complexity in chip manufacturing processes as the drivers behind the strong performance.
Lam Research’s regional revenue breakdown showed China at 35%, South Korea at 20%, and Taiwan at 18%. South Korea and Taiwan combined accounted for 38% of total sales.
This expansion in capital expenditures is mirroring domestic semiconductor giants as well. Samsung Electronics plans to pump more than 110 trillion won into facility investments and R&D this year, while SK hynix intends to ramp up its facility investments from roughly 30 trillion won last year to the upper 40-trillion-won range. Both companies are accelerating efforts to expand production capacity for high-bandwidth memory (HBM) and advanced DRAM.
The earnings of US AI companies demonstrate that revenue growth originating in services and advertising is spreading into demand for servers and semiconductor equipment. Even as debates over capital recovery persist, demand for AI infrastructure continues to underpin capacity expansions by domestic memory makers.
