The Physical Layer

AI and energy infrastructure, explained plainly.

Companies

Nvidia

On the map: Memory · Who runs the capacity · The network between chips · Cooling · Power inside the building · The grid connection · Chips and the factories that make them

Named in 7 pieces. Each entry below is that piece's own description of what the company makes, from its exposure map.

  1. The buyers who design it

    Nvidia and AMD design AI chips; Amazon's Annapurna Labs is the first named NVHBM partner.

    Memory, not maths, sets the pace of AI, so the biggest buyer now designs it · 31 October 2026

  2. The chips

    Nvidia makes the GPUs and rack systems.

    Contracted, energised, running: how Abilene, Nebius and Jane Street's AI capacity really arrives · 25 September 2026

  3. Switches

    Broadcom, Nvidia and Marvell make switch chips. Arista and Cisco build the boxes.

    How ten thousand chips talk: the glass stays, and the plug survives its replacement · 4 September 2026

  4. The halls that choose

    Microsoft, Google, Meta, Amazon and Jane Street run data centres, filled with chips Nvidia and AMD design.

    Heat and water decide where AI goes: liquid to the chip, a choice on the roof · 21 August 2026

  5. At the chip

    Texas Instruments, Monolithic Power, Analog Devices and Renesas make power stages and eFuses. Nvidia sets the 800 VDC design.

    Powering the rack: 800 volts reaches the AI rack before it reaches the building · 7 August 2026

  6. What fills the halls

    Nvidia makes AI chips.

    Johor and Singapore: the plug, not the chip, decides who gets to build · 27 July 2026

  7. The buyer

    Nvidia designs AI chips.

    The chip supply chain's slow parts are sold years before they are made · 14 July 2026

This maps who operates in each layer. It is not a recommendation, and naming a company is not a view on its shares. Personal research, not investment advice.