News
EveryChip

TSMC

Taiwan Semiconductor Manufacturing Company, the largest contract chipmaker. 198 chips on EveryChip so far.

Taiwan Semiconductor Manufacturing Company (TSMC), founded in 1987 by Morris Chang in Hsinchu, is the world's largest contract chipmaker. It makes chips designed by others, including Apple, Nvidia, AMD, Qualcomm and MediaTek, and sells none of its own. TSMC has more than 60 percent of the foundry market and makes most of the world's most advanced chips, at fabs in Hsinchu, Tainan, Taichung and Kaohsiung, with newer fabs in Arizona, Japan and Germany.

The pure-play foundry

When TSMC started, most chip companies ran their own fabs. Chang's idea was a pure-play foundry that would never compete with its customers. That let fabless firms such as Nvidia and Qualcomm grow without building factories costing billions. TSMC took the lead in process technology in the late 2010s with its 7 nm and 5 nm nodes, as Intel's 10 nm was delayed.

Nodes and packaging

TSMC began volume production of its 3 nm N3 process in late 2022 and of its 2 nm N2, its first with gate-all-around transistors, in the last quarter of 2025. Its CoWoS and SoIC packaging join chiplets and HBM memory and are used for most large AI accelerators. TSMC Arizona began volume production in late 2024, with a planned total investment of 165 billion US dollars.

Famous TSMC chips

Further reading

🇺🇸 United StatesAMD Instinct AMD's data centre GPU line for deep learning and HPCGPU · GPU architecture · 2017 WorldGeForce 700 series Nvidia graphics card series from 2013 using Kepler and Maxwell chipsGPU · GPU architecture · 2013 CDNA 🇺🇸 United StatesCDNA AMD microarchitecture for data centre GPU acceleratorsGPU · GPU architecture · 2020 🇺🇸 United StatesHopper Nvidia's datacentre GPU architecture, first shipping in the H100 acceleratorGPU · GPU architecture · 2022 Volta 🇺🇸 United StatesVolta Nvidia's 2017 GPU architecture, first with Tensor Cores for deep learning.GPU · GPU architecture · 2017 Tesla V100 🇺🇸 United StatesTesla V100 Nvidia Volta GPU for AI and HPC data centresGPU · GPU architecture · 2017 AMD Instinct MI250X 🇺🇸 United StatesAMD Instinct MI250X Dual-GPU accelerator for supercomputersGPU · GPU architecture AMD Instinct MI100 🇺🇸 United States and 1 moreAMD Instinct MI100 CDNA architecture GPU for high-performance computing and AI workloads.GPU · GPU architecture · 2020 Intel Data Center GPU 🇺🇸 United StatesIntel Data Center GPU Max Data centre GPU for HPC and AI workloadsGPU · GPU architecture A100 🇺🇸 United StatesA100 Data centre GPU for AI and high-performance computing.GPU · GPU architecture · 2020 GH200 Grace Hopper Sup 🇺🇸 United StatesGH200 Grace Hopper Superchip Combined Arm CPU and Hopper GPU for AI and HPC workloadsCPU · Arm 🇺🇸 United StatesHopper Nvidia datacentre GPU microarchitecture for AI and high-performance computing.GPU · GPU architecture · 2022

Photo credits on this page

The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.

The weekly digest

Every Monday: the chip of the day, new pages on EveryChip, and the conferences and anniversaries of the week ahead.