Skip to content

Ranked list · Infrastructure layer

Companies exposed to Scale-up fabric

The AI stocks the model maps to Scale-up fabric, part of the infrastructure layer, ranked by how strongly each name is exposed.

Ranked by exposure. How strongly the model ties each company to Scale-up fabric, from 0 to 100, with the role it plays there.

16companies14makers

Where it sits

Where Scale-up fabric sits in the chain

What is Scale-up fabric?The very fast links that join many chips into one big tightly-coupled machine.

It defines the size and efficiency of the tightest compute domain. When Scale-up fabric tightens, the companies ranked here are the ones in its path - which is why the model tracks their exposure to it.

Full explainer: what Scale-up fabric is & who makes it
Scale-up fabric · the chain1 of 80 parts
The list, by role
  • Supplier10 companies
  • Producer4 companies
  • Integrator2 companies
The ringed cell is Scale-up fabric, one of 24 parts in the infrastructure layer (tinted).

Questions

Frequently asked

What Scale-up fabric is, who is exposed to it, and what happens when it runs short.

What are the top Scale-up fabric stocks?

By modeled exposure to Scale-up fabric: CoreWeave, Inc., NVIDIA, Huawei, Nebius Group N.V., Amphenol Corporation, Astera Labs. The full ranked list of 16 names is on this page.

Which companies make Scale-up fabric?

The names the model tags as producers or suppliers of Scale-up fabric: NVIDIA, Huawei, Amphenol Corporation, Astera Labs, BizLink Holding Inc., Microchip Technology.

What is Scale-up fabric?

The very fast links that join many chips into one big tightly-coupled machine.

What happens if Scale-up fabric runs short?

Bandwidth or reach limits cap pod size, fragmenting training and lowering efficiency.

See Scale-up fabric in the live chain.

Inside THE ENTITY every part opens on who makes it, who depends on it and where pressure is building, updated nightly.

Exposure scores are illustrative, order-of-magnitude reads from our model of the AI value chain - not investment advice.

as of 2026-11-09Medium confidence model v2.01