Fast peer-cloud growth makes a bullish Amazon earnings view hard to resist. Azure grew 48% and Google Cloud 103%; the forecast is for 42%+ AWS growth. AWS is expected to grow over 40% annually over the next two years, so selling Amazon here risks missing that growth. x.com  x.com 

Amazon is lining up $10 billion more for OpenAI, a company whose CFO Sarah Friar has voiced worry, per Bloomberg, that it may not be able to pay future compute contracts if revenue does not grow fast enough. OpenAI, valued at $852 billion with no actual profit, is a hole that eats Nvidia GPUs and investor cash, and AWS ends up funding a customer it competes against. This looks like a WeWork moment in data centers, not an iPhone moment. x.com 

Everpure, formerly Pure Storage, is a sneaky data-center play that should keep rising steadily. Per @wallstengine, it guides FY28 revenue to $7.0-7.3B, up 39-45% and well above the $6.19B estimate, and newer areas including Scale AI, Modern Data Software and Hyperscale Solutions could reach about 20% of revenue by FY30. x.com 

Amazon has become one of the largest AI chip makers. Its own Trainium and Graviton silicon just crossed a $25bn annualized run rate, up 25% from the prior quarter. Graviton is used by 98% of AWS's top 1,000 customers, the new generation is scaling nearly twice as fast as the last, and Anthropic and OpenAI have multi-gigawatt commitments on Trainium. x.com 

Amazon remains attractive here after announcing a partnership with OpenAI, which came one day after changes to Microsoft's OpenAI partnership. x.com 

Everpure's growth outlook is far stronger than expected. It guided FY28 revenue to $7.0-7.3B, 45% growth and well above the $6.19B consensus. Management also said Scale AI, Modern Data Software and Hyperscale Solutions could reach 20% of revenue by FY30, pointing to an AI-driven opportunity beyond core storage. x.com 

Oracle's OpenAI contracts are valued at very little because the market expects them never to happen, so they should reprice as OpenAI proves fine. OpenAI is the only lab running both AI revenue engines: enterprise subscriptions and API fees, which carried Anthropic past it in ARR, plus over a billion mostly free ChatGPT users now being monetized through ads, a business that hit a $1 billion run rate six months after launch. Both labs will also need far more compute. x.com 

Bloom Energy's valuation is silly because it still has no real earnings despite great results. It has no GAAP earnings and no free cash flow excluding stock-based compensation. Demand is off the charts, Oracle validated its technology and it seems to be in pole position, but the economics are unimpressive given all that demand. x.com  x.com  x.com  x.com  x.com 

Everpure, formerly Pure Storage, is a key player in the coming flash choke point, and the stock should start to fly once Google Genie comes back. @itschrisray sees it as a huge beneficiary of the AI buildout through memory efficiency: proprietary DirectFlash Modules and Purity OS compression let drives store up to 5x more data with 50-80% less power, keeping Nvidia GPUs fed. x.com 

Oracle faces double jeopardy, viewed both as a legacy software business and as an overlevered infrastructure business that depends on OpenAI being good for its capex commitments. OpenAI is already so aggressively valued that even catching Anthropic in private valuation leaves limited upside, which also limits how much SoftBank, which @goodalexander calls leveraged long OpenAI, could benefit. x.com