SpaceX reported fiscal Q2 2026 revenue of $7.81 billion, doubling from $4.07 billion a year earlier. Starlink subscribers reached 12.0 million, up from 6.0 million, while the company spent $15.83 billion on AI in the quarter. SpaceX also agreed to acquire AI coding firm Cursor for $60 billion, though it posted a net loss of $541 million.
Cerebras saw core revenue jump 103% year over year to $209.9 million in Q2, with cloud and services revenue climbing 281% to $126.0 million. Customers include OpenAI, Block, Figma, and AWS. The company carries a $25.4 billion backlog and over 600 megawatts of data center capacity under contract, but posted a GAAP net loss of $450.5 million.
Nebius raised GPU prices on its NVIDIA B300, B200, H200, and H100 virtual machines by more than 30%. Revenue hit $582.3 million, up 454% year over year, with $3.0 billion in annualized recurring revenue. Microsoft and Meta are among clients with over $40 billion in commitments, though the company spent $8.1 billion on equipment and carries $8.5 billion in debt.
Synopsys reported fiscal Q3 revenue of $2.477 billion and raised full-year targets to $9.715 billion. At its Investor Day, management guided fiscal 2027 revenue to $11.1 billion to $11.2 billion, announcing a multi-year IP deal with Amazon and a partnership with OpenAI to build GPT-Synopsys for chip design.
Teradyne posted record Q2 revenue of $1.33 billion, up 104% from $652 million a year earlier, driven by memory chip testing demand. The company also demonstrated physical AI robots using its Universal Robots hardware at the Automate trade show, though it guided Q3 revenue lower to $1.2 billion to $1.3 billion.
Reach Security raised over $50 million in Series B funding led by Top Tier Capital Partners, with participation from Ballistic Ventures, In-Q-Tel, M12, and Microsoft's Venture Fund. The company launched Reach Everywhere and uses AI to review security controls at machine speed. AI-enabled adversary operations have increased 89% and breakout times have dropped to as little as 27 seconds.
Anthropic plans to share updated financial numbers with institutional investors this week as it prepares for what could be the largest IPO in history. The company may value itself at up to $2 trillion and filed confidentially for an IPO earlier this year without releasing its S-1 forms publicly.
The IMF warned of high concentration risks in AI hardware and markets, noting that global economic recovery is narrow and driven mostly by an AI investment boom in select advanced economies and Asia. Fiscal deficits are growing and borrowing costs are rising in advanced markets.
Massachusetts is close to passing one of the strictest AI safety laws in the United States. Governor Maura Healey supports setting reasonable guardrails, with independent evaluators testing AI models for catastrophic risk at least three times a year. Google, Anthropic, OpenAI, and Meta are lobbying the legislature, though Anthropic and OpenAI both support independent evaluations.
Australia may increase funding for national AI security. Over 300 scientists signed an open letter calling for coordinated efforts against rogue AI agents. Industry Minister Tim Ayres said the government wants more control over AI models by training them locally, though he did not explain how copyright rules would change to allow this.
Key Takeaways
- SpaceX Q2 revenue doubled to $7.81 billion and Starlink hit 12.0 million subscribers, but the company also agreed to buy Cursor for $60 billion while posting a $541 million net loss.
- Cerebras revenue grew 103% year over year to $209.9 million with a $25.4 billion backlog, though it posted a $450.5 million GAAP net loss and trades at 217.39 times forward earnings.
- Nebius raised GPU prices by over 30% on NVIDIA B300, B200, H200, and H100 VMs, with $40 billion in customer commitments from Microsoft and Meta but $8.5 billion in debt.
- Synopsys guided fiscal 2027 revenue to $11.1 billion to $11.2 billion and announced a multi-year IP deal with Amazon plus an OpenAI partnership for GPT-Synopsys chip design.
- Teradyne reported record Q2 revenue of $1.33 billion, up 104% year over year, but guided Q3 lower to $1.2 billion to $1.3 billion.
- Reach Security raised over $50 million in Series B funding including participation from Microsoft's Venture Fund, as AI-enabled adversary operations increased 89% and breakout times dropped to 27 seconds.
- Anthropic may value itself at up to $2 trillion as it prepares for what could be the largest IPO in history, having filed confidentially earlier this year.
- The IMF warns that global economic recovery is narrow, driven mostly by an AI investment boom in select advanced economies and Asia, with rising fiscal deficits and borrowing costs.
- Massachusetts is close to passing strict AI safety legislation requiring independent evaluators to test models for catastrophic risk at least three times a year, with Google, Anthropic, OpenAI, and Meta lobbying.
- Australia may boost AI security funding as over 300 scientists call for global oversight, with the government seeking more local control over AI model training.
SpaceX leans on Starlink as AI spending soars
SpaceX reported fiscal Q2 2026 revenue of $7.81 billion, up from $4.07 billion a year earlier. Starlink reached 12.0 million subscribers, up from 6.0 million a year earlier, while AI spending hit $15.83 billion in the quarter. The company posted a net loss of $541 million and agreed to buy AI coding firm Cursor for $60 billion. Investors are concerned because capital spending far outpaces operating cash flow, and the stock trades at 200 times forward earnings.
Cerebras bets on cloud growth despite steep valuation
Cerebras reported second-quarter core revenue of $209.9 million, up 103% year over year, with cloud and services revenue jumping 281% to $126.0 million. The company holds a $25.4 billion backlog and has over 600 megawatts of data center capacity under contract. OpenAI, Block, Figma, and AWS are among its major customers. However, the stock trades at 217.39 times forward earnings, and the company posted a GAAP net loss of $450.5 million in the quarter.
Nebius raises GPU prices as debt-heavy buildout continues
Nebius raised prices on its NVIDIA B300, B200, H200, and H100 virtual machines by more than 30% from the first quarter. Second-quarter revenue reached $582.3 million, up 454% year over year, with annualized recurring revenue hitting $3.0 billion. The company has over $40 billion in customer commitments from firms like Microsoft and Meta. However, it spent $8.1 billion on equipment in the first half of 2026, carries $8.5 billion in debt, and posted a GAAP net loss of $190.4 million.
Synopsys unveils $11B growth plan after Investor Day
Synopsys reported fiscal Q3 revenue of $2.477 billion, up from $1.740 billion a year earlier, and raised full-year targets to $9.715 billion in revenue. At its Investor Day, management guided fiscal 2027 revenue to $11.1 billion to $11.2 billion and announced a multi-year IP deal with Amazon and a partnership with OpenAI to build GPT-Synopsys for chip design. The stock trades at 33.69 times forward earnings, and the company expects about $3.1 billion in free cash flow for fiscal 2027.
Teradyne rides AI chip testing boom to record revenue
Teradyne reported record second-quarter revenue of $1.33 billion, up 104% from $652 million a year earlier, driven by strong demand for memory chip testing. The company also showed physical AI robots at the Automate trade show using its Universal Robots hardware. However, it guided third-quarter revenue to $1.2 billion to $1.3 billion, below the record quarter. The stock trades at 44.29 times forward earnings, and short interest stands at 6.38% of the float.
Reach Security raises over $50M to fight AI-powered cyber attacks
Reach Security secured more than US$50 million in Series B funding led by Top Tier Capital Partners. The company launched Reach Everywhere, a tool that brings security control insights to wherever defense teams work. Reach uses AI to review and tune security controls at machine speed, closing gaps that AI-powered attackers exploit. The company has doubled its customer base in 12 months, with clients including Autodesk and Ensemble Health Partners.
Reach Security secures $50M to outpace AI cyber threats
Reach Security raised over $50 million in a Series B round with participation from Ballistic Ventures, In-Q-Tel, M12, and Microsoft's Venture Fund. The funding will support product innovation, integration partnerships, international expansion, and growth in US enterprise and federal markets. Reach uses AI to autonomously review every security control and close entry points used by attackers. The company reports that AI-enabled adversary operations have increased 89% and breakout times have dropped to as little as 27 seconds.
High-stakes earnings season hits expensive AI stocks
AI stocks like AMD, Marvell, Corning, and Bloom Energy saw sharp valuation declines according to Trivariate Research. The most expensive stocks experienced multiple contractions roughly 100 times larger than the cheapest names. The S&P 500's forward price-to-earnings ratio fell from 22 at the start of 2026 to 19 by late last week. Rising US Treasury yields to multidecade highs are pressuring investors to demand stronger proof of future growth from high-priced names.
ChatPlayground offers ChatGPT, Gemini, Claude for $60 lifetime
ChatPlayground AI brings ChatGPT, Claude, Gemini, DeepSeek, Llama, and Perplexity together in one interface. Users can enter a prompt once and compare responses side by side. The Unlimited Plan includes unlimited monthly messages, priority support, and early access to new features. The lifetime subscription is priced at $60. Users can also upload PDFs and images to ask questions about their files.
How to fix what ChatGPT and Gemini say about you
AI tools like ChatGPT and Gemini create summaries about people by pulling from the open web, not from editable profiles. ChatGPT reached 900 million weekly active users in February 2026, and Gemini passed 750 million monthly users by late 2025. Reputation Resolutions recommends running an AI reputation audit by searching questions like Who is [name] across all major platforms and saving screenshots of every answer. The key insight is that fixing the source pages is what changes the AI answers, since every response traces back to something on the open web.
Anthropic to reveal new financial details ahead of possible IPO
Anthropic plans to share updated financial numbers with institutional investors this week. The company is preparing for what could be the largest IPO in history. Revenue numbers from private AI companies have recently caused big market swings. Anthropic may value itself at up to 2 trillion dollars. The company filed confidentially for an IPO earlier this year but has not released its S-1 forms publicly.
IMF warns of high concentration risks in AI hardware and markets
The International Monetary Fund says global economic growth is showing resilience despite trade and commodity shocks. However, the recovery is narrow and driven mostly by an AI investment boom in select advanced economies and Asia. Fiscal deficits are growing and borrowing costs are rising in advanced markets. Emerging economies and African nations face different paths forward.
Big 1990s tech shareholder sees new AI upside in Flex
Paul Meeks was one of the biggest shareholders in a major tech stock during the 1990s. He is bullish on Flex again because of AI. Flex plans to spin off Axiom, its AI infrastructure business, in 2027. Meeks sees potential upside as data center power demand continues to grow.
Massachusetts may pass country's strictest AI safety law
Massachusetts is close to passing one of the strictest AI safety laws in the United States. Governor Maura Healey supports setting reasonable guardrails for AI companies. The proposed law would use independent evaluators to test AI models for catastrophic risk at least three times a year. Companies like Google, Anthropic, OpenAI and Meta are lobbying the legislature. Anthropic and OpenAI both support independent evaluations of their AI technology.
Australia seeks AI security boost as scientists call for global oversight
The Albanese government in Australia may increase funding for national AI security. Over 300 scientists and researchers signed an open letter calling for coordinated efforts against rogue AI agents. Industry Minister Tim Ayres said the government wants Australia to have more control over AI models by training them locally. The minister did not explain how copyright rules would change to allow this training.
Sources
- SpaceX (SPCX): Can Starlink Carry the AI Bill?
- Can Cerebras (CBRS) Turn Its $25.4B Backlog Into High-Margin Growth?
- Nebius (NBIS) Has Pricing Power. Its Earnings Don’t Show It Yet.
- Synopsys (SNPS) Just Laid Out a $11B AI Roadmap. Is the Stock a Buy Right Now?
- AI Chips Need Testing, and Teradyne (TER) Is Winning. Is the Valuation Too High?
- Reach Security raises more than US$50M Series B to stop AI-Powered attacks by autonomously closing the gaps they target
- Reach Security Raises More Than $50M Series B to Stop AI-Powered Attacks by Autonomously Closing the Gaps They Target
- It’s a high-stakes earnings season, especially for AI stocks
- There’s just a few more days to get ChatGPT, Gemini, Claude, and more for just $60 for life
- How to fix what ChatGPT and Gemini say about you
- Anthropic to disclose new revenue numbers as AI lab financials become central to tech trade
- IMF warns of high concentration risks in ai hardware and markets
- I was one of the biggest shareholders in this tech stock in the 1990s. I'm bullish again because of AI
- Massachusetts May Pass Country’s Strictest AI Safety Law
- Albanese government seeks AI security boost as scientists call for global oversight
Comments
Please log in to post a comment.