Videos
Stories of the hardware behind AI, from the Silicon Ages channel. Lecture playlists are on the way.
Videos
They Built a Chip the Size of a Dinner Plate. Here's How It Works.
Related course: Computer Architecture and AI Hardware
Why One AI Chip Needs a Skyscraper of Memory (HBM Explained)
Related course: Custom Layout and Memory
Intel's $140 Billion Day: The Chip Story Behind the Headline
Related course: Fabrication, Packaging and Careers
The Wires Inside AI Are Hitting a Wall. Light Is the Fix
Related course: Computer Architecture and AI Hardware
Inside an NVIDIA AI Data Center: What Breaks First?
Related course: Computer Architecture and AI Hardware
What Happens After You Press Enter? One AI Prompt, Followed to the Transistor
Related course: Computer Architecture and AI Hardware
Run AI on Your Own Laptop: What Really Happens Inside (Down to One Transistor)
Related course: Computer Architecture and AI Hardware
Why Not Just Use GPUs? Inside TPUs, Custom Chips, FPGAs and Wafer-Scale AI
Related course: Computer Architecture and AI Hardware
Shorts
A 16 GB AI becomes 5 GB with one trick
Your laptop can run an AI with the Wi-Fi off
One AI answer uses about as much energy as 9 seconds of TV
Same question, different answers. Why?
160 stops for every word, down to one transistor
608 billion tokens in one day
You share the bus with strangers
One chat can fill half a GPU: the KV cache
The busiest chip in the world is idle 99% of the time
The AI never sees your words
A 671 GB model does not fit in eight of the best GPUs
New York to London: 55 ms at the very best
Light is slower in glass than in empty space
What happens after you press Enter on an AI prompt?