Explained
One bottleneck at a time: what it is, who pays, and where money sticks in the AI stack.
Explained
GPU as a service is billed accelerator time: a cloud operator puts GPUs in a power-ready hall and sells hours — on-demand, reserved, or as a dated capacity block — instead of the customer buying the cards.
Explained
A neocloud is a specialist cloud that leases GPU clusters for AI work. What it is, how it earns, and why utilisation and realised price decide everything.
Explained
CoWoS is chip-on-wafer-on-substrate — TSMC’s packaging family that joins a compute die and HBM so an AI GPU can ship. How variants differ, and where money sticks when slots bind.
Explained
Behind-the-meter power is on-site electricity on the customer's side of the utility meter. Leased megawatts do not pay until the hall is energised — and the queue is why that gap exists.
Explained
Advanced packaging still needs substrates and ABF film. The quiet neighbour gate — and how inventory builds when materials lag.
Explained
Process tools set HBM bit growth months before stacks ship. The equipment gate, the install-to-output clock, and where cash sits on the supply curve.
Explained
AI chips finish when foundry, packaging, HBM, substrates, and test all match. The wider finish line beyond one packaging brand name.
Explained
Software turns expensive compute into paid workflows. Where app-layer rent sits — and why it runs on a different clock from fabs.
Explained
Servers turn chips, networking, and power into a deployable rack. This Map explains what fills the box — and when system lead times become the gate.
Explained
Speed-to-power is the clock from plan to live watts at the rack. Chip news alone misses this gate — and who gets paid when timed megawatts clear.
Explained
Data-centre REITs are landlords in the AI stack. They earn when timed power and long leases meet dense GPU demand — not from empty shells.
Explained
Optical networking is the bandwidth layer of an AI cluster. When fibre, modules, and switches lag GPU counts, finished chips sit underused.
A Mechanism Explained page teaches one stuck step in the AI stack. It names the part, the buyer, and the cash path — then links to siblings, Wire, and Analysis.
URL shape: /explained/<term>-explained/. Length target about 1,500–2,000 words. This hub lists those pages.
Search demand clusters around plain “what is” questions. Use these doors first:
Browse by layer on The Stack. Compare two terms in Versus. Roles only in Players.