A broad rally in U.S. stocks gave way to a far more cautious tone after the close, when SpaceX and Advanced Micro Devices, or AMD, released their latest quarterly results. Despite headline numbers that beat expectations in key areas, both stocks fell sharply in after-hours trading, with SpaceX down more than 7% and AMD off more than 9%. The article argues that the market is no longer rewarding companies simply for spending aggressively on AI or posting solid growth. In SpaceX’s case, investors focused on heavy capital expenditures, including roughly $16 billion for AI computing infrastructure, a figure well above prior analyst expectations. AMD, by contrast, delivered strong revenue, earnings and guidance, yet still sold off as elevated valuations appeared to leave little room for anything short of a much larger beat. The piece frames both reactions as evidence of a shift in how AI-related earnings are judged. According to the original article, companies now need to show disciplined spending, protect their core businesses, and offer very strong forward guidance at the same time. It also includes a risk disclosure stating that the analysis reflects market observation rather than investment advice.

U.S. stocks rallied sharply Monday as strong corporate earnings and signs of easing tension around the Strait of Hormuz lifted risk appetite across markets. The Dow Jones Industrial Average jumped 1.71% to close above 54,000 for the first time, while the S&P 500 rose 1.79% and the Nasdaq Composite gained 2.59%, both setting fresh records. Oil prices extended losses after Axios reported that the U.S., Iran and Oman were nearing a temporary agreement aimed at reopening the Strait of Hormuz, restoring a ceasefire and creating room for follow-up nuclear talks. WTI crude fell 5.58%, and Brent dropped below $80. The move in energy also fed into rates. The 10-year U.S. Treasury yield slipped about 6 basis points to around 4.6%, and CME FedWatch showed the probability of a 25-basis-point rate hike in September falling to 57.1% from 67.2% a day earlier. Investors are now looking to Friday’s nonfarm payrolls report for the next major macro signal. AI remained the strongest equity theme, but this time the rally spread well beyond chipmakers. Semiconductor, memory, optical communications, AI software, and data center infrastructure names all advanced. Palantir, ARM, Intel, Sandisk, Micron, AMD, Nvidia, and Bitdeer were among the names in focus, while markets also turned to upcoming earnings, SpaceX lockup expirations, and storage-sector results later this week.

Advanced Micro Devices reported a sharp jump in data center revenue in the second quarter, with the segment bringing in $6.718 billion, up 107% from a year earlier and accounting for 58% of total company revenue. Total revenue reached $11.536 billion, while non-GAAP earnings per share came in at $1.66, both record highs for the company. AMD also issued third-quarter guidance of about $13 billion in revenue, plus or minus $300 million, and projected a non-GAAP gross margin of roughly 56%, ahead of the average estimates cited by Bloomberg and LSEG. CEO Lisa Su went beyond the quarterly outlook and said on the earnings call that AMD expects data center revenue to more than double again by 2027, adding that the increase could be “well over 100%.” She also signaled that AMD could outperform its prior long-term targets, including 35% annual revenue growth and $20 in earnings per share, and said server revenue growth in the second half of 2026 would exceed 80% year over year. Even so, AMD shares fell nearly 9% in after-hours trading after the results. The report said some analysts had been looking for an even stronger third-quarter outlook, with expectations in some cases as high as $14 billion, helping explain the negative market reaction despite strong fundamentals.

AMD reported Q2 2026 results on August 5, with data center revenue hitting $6.72 billion, up 107% year over year. Total revenue reached $11.54 billion, up 50% and beating estimates. AMD guided Q3 revenue to about $13 billion, above consensus. CEO Lisa Su said second-gen Helios AI servers have entered full production, with shipments to Meta and OpenAI expected by quarter-end. Despite the strong report, shares fell over 9% after hours.

Advanced Micro Devices reported stronger-than-expected results for the second quarter of 2026, posting $11.5 billion in revenue and non-GAAP earnings per share of $1.66, both above Wall Street estimates. The company’s guidance for the third quarter, however, became the market’s main focus. AMD projected midpoint revenue of $13 billion, plus or minus $300 million. While that topped the consensus estimate of $12.5 billion, it fell short of the more aggressive forecasts held by some investors and institutions, sending the stock down about 9% in after-hours trading. The quarter was driven by continued strength in data center chips. Revenue from the data center segment rose more than 100% year over year to $6.7 billion, ahead of the expected $6.6 billion. PC and gaming-related revenue increased 6% to $3.8 billion. On the earnings call, CEO Lisa Su said AMD expects data center revenue to grow by well over 100% in 2027 and reiterated long-term operating growth targets above its earlier benchmarks of 35% annual revenue growth and $20 in earnings per share. ABMedia also noted that memory shortages tied to AI infrastructure demand could still pressure the broader PC market.

Advanced Micro Devices used its Advancing AI 2026 event in San Francisco on July 22-23 to roll out three major products at once: the Instinct MI400 accelerator family, the EPYC "Venice" server CPU, and the Helios rack-scale AI platform. The release, as described by MSX Research Institute, shows AMD shifting its pitch from being an alternative GPU supplier to Nvidia toward becoming a full-stack AI infrastructure provider spanning chips, systems, networking and software. CEO Lisa Su also laid out a broader market thesis, saying the total computing market could reach $2 trillion by 2030, including $1.4 trillion for AI accelerators and $200 billion for data-center CPUs. AMD claimed Helios, built around MI455X accelerators, can outperform Nvidia’s next-generation Vera Rubin platform on some key metrics, while Venice can deliver up to 3.4x the performance of Intel Xeon and about 20% more than Nvidia’s Vera CPU. On the customer side, Meta, OpenAI and Anthropic have all committed gigawatt-scale orders. MSX said Meta and OpenAI each secured warrants for as many as 160 million AMD shares, equivalent to about 10% of AMD equity each, while AMD may invest up to $5 billion in Anthropic. Still, MSX noted the performance claims are AMD’s own and said investors are now watching whether MI400, Venice and Helios can turn into sustained revenue.

Intel reported revenue of $16.1 billion for the second quarter of 2026, up 25% year over year and ahead of the $14.42 billion analyst consensus tracked by LSEG. On a non-GAAP basis, the company posted net income of $2.2 billion, or $0.42 per share, compared with a loss a year earlier and above the $0.21 per-share estimate. The standout segment was Data Center and AI, which generated $6.3 billion in revenue, up 59% from a year ago, as CEO Lip-Bu Tan said CPU demand in data centers is "taking off" and running ahead of supply. Intel also said it has signed 10 long-term agreements of that kind while operating under supply constraints. Foundry revenue rose 31% to $5.8 billion, though the unit remained in an operating loss due in part to internal manufacturing activity. Intel also disclosed Fortinet as the first publicly named customer for its foundry business. Looking ahead, the company raised its 2026 capital spending plan to $20 billion from $18 billion and issued third-quarter guidance above Wall Street expectations, including revenue of $15.8 billion to $16.8 billion, non-GAAP EPS of $0.38, and gross margin of about 42%.

AMD used its Advancing AI event in San Francisco, held July 22-23, to lay out the next phase of its AI hardware push. CEO Lisa Su said the company’s rack-scale AI system Helios has entered full production and will begin shipping soon. OpenAI’s infrastructure chief said the company expects to deploy AMD Helios at scale, while OpenAI also plans to work with AMD on the MI500 series of AI chips and later products. Su also said AMD is working with chip design company Cerebras to deliver high-speed inference through Cerebras’ cloud service. That joint offering is expected to come to market later this year. AMD and Cerebras plan to combine Helios GPU server racks with Cerebras wafer-scale chips for AI inference. CNBC noted that Su had projected the AI accelerator market at $500 billion by 2028 a year ago, and the latest outlook now puts the market by the end of this decade at roughly the size of today’s entire semiconductor industry, with GPUs expected to account for most of it.