Who makes
AI hardware.

Designer of Trainium and Inferentia accelerators and the largest cloud operator.

Amazon designs Trainium (training) and Inferentia (inference) chips to lower its reliance on merchant GPUs, and operates AI infrastructure at hyperscale through AWS.

GPUs and AI accelerators

Designer of Instinct data-center GPUs, the main merchant alternative to NVIDIA for AI compute.

AMD's Instinct line and its ROCm software are the leading challenger to NVIDIA in data-center AI. Like NVIDIA it is fabless and relies on TSMC and on HBM suppliers.

GPUs and AI accelerators

Designer of TPU accelerators and a hyperscale AI data-center operator.

Google designs its own TPUs to train and serve its models and runs large AI data centers, making it both a chip designer and an infrastructure operator.

GPUs and AI accelerators

Dominant designer of AI training and inference GPUs, and of the CUDA software stack and NVLink/networking around them.

NVIDIA designs the accelerators most frontier AI is trained on and increasingly sells whole rack-scale systems. Its moat is as much CUDA and the surrounding ecosystem as the silicon. It is a fabless designer: TSMC manufactures the chips.

GPUs and AI accelerators

The third major HBM supplier and a large memory maker.

Micron rounds out the trio of HBM suppliers and is the main US-based memory producer, giving it strategic importance amid supply-chain reshoring.

High-bandwidth memory (HBM)

Major HBM and DRAM supplier and an alternative leading-edge foundry.

Samsung is one of three HBM makers and also runs a leading-edge foundry, giving it a stake in two of the tightest links in the chain.

High-bandwidth memory (HBM)

Leading supplier of high-bandwidth memory for AI accelerators.

SK Hynix has led the HBM market that feeds AI GPUs. Because only three firms make HBM at scale, its output materially affects finished accelerator supply.

High-bandwidth memory (HBM)

Integrated device maker pursuing a leading-edge foundry business and AI accelerators.

Intel is building out Intel Foundry to offer leading-edge US manufacturing and develops its own accelerators, positioning itself as a domestic alternative across several layers.

Chip foundries and fabrication

The leading-edge foundry that manufactures the large majority of advanced AI accelerators and owns CoWoS advanced packaging.

TSMC is the single most important manufacturer in AI hardware: it fabricates leading-edge chips for NVIDIA, AMD, Apple and others, and its CoWoS packaging is a recurring bottleneck. Its concentration in Taiwan is a central geopolitical fact of the industry.

Chip foundries and fabrication

Sole maker of EUV lithography machines required to fabricate leading-edge chips.

ASML's EUV scanners are a single-supplier dependency for the entire leading-edge industry, which makes the company both indispensable and a focal point of export controls.

Lithography and fab equipment

Maker of the precision optics inside EUV lithography systems.

Zeiss SMT supplies the extreme-precision mirrors and optics that make EUV lithography possible, an upstream single-source dependency behind ASML itself.

Lithography and fab equipment

Supplier of networking switch silicon and a partner for custom AI accelerators (ASICs).

Broadcom supplies much of the switch silicon that knits AI clusters together and co-designs custom accelerators with hyperscalers, making it central to both networking and bespoke compute.

Networking and interconnect

Contract manufacturer assembling AI servers and systems at very large scale.

Foxconn (Hon Hai) is one of the largest contract manufacturers building AI servers for cloud providers and OEMs, a key link between chip suppliers and deployed systems.

AI servers and systems

Major integrator of dense, liquid-cooled AI server systems.

Supermicro assembles accelerators, networking, power and cooling into AI server systems at scale, and has been an early mover in factory-scale liquid-cooled deployments.

AI servers and systems