Everything I write about artificial intelligence lands here. Less of the demo-day theater, more of what sits underneath it: the compute, the real economics, and the distance between what gets announced and what actually ships.
The ASML lithography monopoly is the most complete chokepoint in tech. High-NA in fabs, Hyper-NA ahead, China capped, one company holding it all up.…
Snapdragon Insider bias declared. Qualcomm's AI200 data center card plus the five-form-factor Hexagon NPU strategy is a bet that inference is where the real…
TSMC's A16 node brings backside power delivery (Super Power Rail) in 2026, answering Intel's 18A and opening the angstrom era for AI and HPC…
LeCun says xAI is 'kind of a failure.' With a $2.5B quarterly loss and compute rented to rivals, the numbers are hard to dismiss.
Midjourney's full-body ultrasound scanner promises 60-second scans. The prototype takes 20 minutes and has scanned just 12 people.
The Fable 5 shutdown highlighted the importance of owning AI models on personal hardware to avoid government restrictions. Although open-source AI is gaining traction,…
Qualcomm's Snapdragon Reality Elite puts the NPU first: 48 TOPS, Android XR, and the XREAL Aura. Here's what the chip actually changes for XR…
Sony's L910 sensor hits 100 dB HDR in one shot using LOFIC and TCG-HDR—eliminating motion artifacts that plagued multi-frame rivals.
NVIDIA's data center revenue surpassed $100 billion in fiscal 2026, bolstered by innovations like the Blackwell B200, which significantly enhances performance. The acquisition of…
Qualcomm's Snapdragon Reality Elite lands 48 TOPS of on-device AI compute in XR headsets, a 160% NPU jump over the XR2+ Gen 2 that…
The EU AI Act highlights a disconnect between regulatory ambitions and Europe's hardware capabilities, creating compliance burdens for startups while relying on US cloud…
SpaceX went public without loosening Musk's grip. The US government picked quantum winners with political ties. AI agents started trading crypto alone. And the…
Satya Nadella's recent LinkedIn piece outlines a framework for companies to build both human and token capital in AI. He emphasizes the need for…
Reasoning models are measurably smarter. The real question is who controls the infrastructure they run on, and whether better reasoning creates value or creates…
128 GB of unified memory, a 40-core RDNA GPU, and local LLM inference nobody can revoke. The GMKtec EVO-X2 is the hardware argument for…
One maker's tech digest, weekly. Chips, AI, gaming, EVs — no spam.