#AI Chips

SpaceX Colossus 2 AI data center exterior in Memphis Tennessee, housing Nvidia GB300 GPU clusters for xAI and commercial compute clientsStartups

Reflection AI Is Paying SpaceX $150 Million a Month. No Model Has Shipped.

Reflection AI started paying SpaceX $150 million a month on July 1 for Nvidia GB300 access at Colossus 2. No model has shipped yet.

Saganote•
Illustration showing Meta's AI cloud computing initiative and its competitive impact on CoreWeave's AI infrastructure businessTech

Meta Is Building a Cloud Business to Sell AI Compute, and CoreWeave Just Took the Hit

Bloomberg reported July 1 that Meta plans a cloud business to sell surplus AI compute. Meta stock jumped 10%. CoreWeave dropped 14%. Nebius 16%.

Saganote•
Apple MacBook Pro on a desk representing the upcoming Apple M7 chip generation and its AI-focused Apple silicon roadmapAI

Apple Skips M6 Pro and Max to Fast-Track an AI-Focused M7 Chip for 2027

Apple is skipping the M6 Pro and Max to launch an AI-focused M7 chip lineup in 2027, Bloomberg reports. On-device AI inference drove the decision.

Saganote•
Micron Hbm Memory Chip Semiconductor Wafer 2026AI

Micron Revenue Quadrupled to $41 Billion. Wall Street Is Calling It the Next Nvidia.

Micron revenue quadrupled to $41 billion last quarter as AI data centers ran out of HBM. Wall Street briefly pushed Micron past Meta's market cap and is calling it the next Nvidia.

Saganote•
Google Search homepage displayed on a smartphone in dark mode with the Google logo blurred in the backgroundAI

Google Cuts Off Meta's Gemini Access as AI Compute Capacity Runs Short

Google has restricted Meta's access to its Gemini AI models after the social media giant sought more compute capacity than Google could provide, disrupting internal AI projects.

Saganote•
Microsoft Maia 200 custom AI accelerator chip on a circuit board, announced January 2026AI

Anthropic in Talks to Run Claude on Microsoft's Maia 200 Chip

Anthropic is in talks to run Claude on Microsoft's Maia 200. A deal would make it the first major external workload on Microsoft's custom AI chip.

Saganote•
OpenAI President Greg Brockman and Broadcom CEO Hock Tan holding the Jalapeño custom AI inference chip at its official unveiling in June 2026AI

OpenAI and Broadcom Unveil Jalapeño: First Custom Chip Targets LLM Inference

OpenAI and Broadcom unveiled Jalapeño, a custom AI inference chip for LLMs, built in nine months with substantially better performance per watt.

Saganote•