Aug 30, 2026
•
3 min read
Why buying a model hub might matter more than buying another chip fab
Aug 29, 2026
4 min read
Thinking through how NVIDIA's healthcare stack could map onto triage and why the human has to stay the one making the final call.
Aug 28, 2026
A look at why nuclear power is entering the picture, what's actually cooling these racks, and the real story behind AI's water use.
Aug 13, 2026
A look at the power, cooling, and site-selection differences behind AI data centers and whether traditional facilities are actually going away.
Aug 12, 2026
A look at why agentic and physical AI need so much more compute, who's building the infrastructure for it, and whether the spending matches the demand.
Aug 11, 2026
A look at compliance, latency, and cost and the order they tend to apply in.
Aug 10, 2026
A quick look at the chips, connections, and training steps behind every AI model GPU, TPU, NPU, and how it all fits together.
Aug 4, 2026
A closer look at how compute, network, and storage work together and why the slowest one decides everything.
Aug 3, 2026
Let's figure out if this is a GPU problem or a 'pick the right AI model' problem
