COMPANIES · CHIP MFG & PACKAGING

TSMC

Manufactures semiconductors designed by customers and provides advanced packaging services.

EnablerChip mfg & packaging
What it does

Manufactures semiconductors designed by customers and provides advanced packaging services.

Why it matters

It is the primary manufacturer of the advanced logic chips (including NVIDIA's and most other AI accelerators) that the entire AI hardware stack depends on.

How it makes money

Manufactures chips designed by other companies (a 'foundry') and charges for wafer fabrication and packaging services.

Geography

Taiwan

AI stack position

Chip manufacturing, memory and packaging. The foundries, memory makers, and packaging and testing businesses that physically manufacture AI chips. This is where a chip design actually becomes a real, physical piece of silicon.

Connections (21)

Every connection below is tied to a readable source, date, classification and statement of its limits. A company source establishes what that company stated; it does not necessarily provide independent confirmation.

Current relationships (12)

TSMC manufactures AI accelerators and data-center processors for AMD
Manufactures forCurrent
TSMC manufactures AMD's MI-series AI accelerators and data-center processors; press analysis of TSMC's July 2026 revenue (up 45% year-over-year) cites AMD's AI-chip demand as a primary driver of TSMC's high-performance-computing revenue, which reached 66% of quarterly sales.

Source: 2026-08-12 · reviewed 2026-09-22

Evidence: Secondary reporting (unconfirmed by primary parties) · Reported operating

Limits: Press reporting, not a joint statement from TSMC or AMD naming the arrangement; discloses no contract terms, wafer volumes, or exclusivity.

TSMC Revenue Surges 45% as AI Chip Demand From Nvidia, AMD, and Apple ↗
TSMC manufactures custom AI ASICs for Broadcom
Manufactures forCurrent
TSMC manufactures Broadcom's latest custom AI ASICs on its 2-nanometer process using 3.5D advanced packaging, according to trade-press reporting on Broadcom's custom-silicon roadmap.

Source: 2026-03-05 · reviewed 2026-09-22

Evidence: Secondary reporting (unconfirmed by primary parties) · Reported operating

Limits: Trade-press reporting rather than a direct statement from TSMC or Broadcom; no volumes, contract value, or product names are disclosed.

Broadcom-TSMC 3.5D AI chips give ASIC leader an early edge over Nvidia ↗
TSMC manufactures chips for Microsoft
Manufactures forCurrent
Industry reporting indicates Microsoft's Maia 200 AI accelerator is built on TSMC's 3nm process node.

Source: 2026-01-27 · reviewed 2026-09-22

Evidence: Secondary reporting (unconfirmed by primary parties) · Reported operating

Limits: Does not confirm order volumes or exclusivity; TSMC does not comment on individual customers, and Microsoft's own materials do not name the foundry.

Microsoft Unveils Maia 200 AI Chip on TSMC 3nm; SK hynix Reportedly Sole HBM3E Supplier ↗
TSMC manufactures chips for Google (Alphabet subsidiary)
Manufactures forCurrent
Industry reporting indicates Google's first Arm-based CPU, Axion, is built on TSMC's 3nm process with ASIC design support from GUC (Global Unichip).

Source: 2025-10-21 · reviewed 2026-09-22

Evidence: Secondary reporting (unconfirmed by primary parties) · Reported operating

Limits: Does not confirm volumes or exclusivity; TSMC does not comment on individual customers and this is based on supply-chain analyst reporting.

Google's First Arm-Based CPU Axion Reportedly Built on TSMC 3nm with GUC Design Support ↗
TSMC manufactures chips for NVIDIA
Manufactures forCurrent
TSMC began Blackwell production in Arizona.

Source: 2025-04-14 · reviewed 2026-09-17

Evidence: Primary-source statement · Reported operating

Limits: Does not establish volumes, pricing or exclusivity.

NVIDIA US manufacturing announcement ↗
TSMC is fabricating a custom chip for AST SpaceMobile
Manufactures forCurrent
AST SpaceMobile said its AST5000 ASIC, designed to increase per-satellite processing bandwidth roughly tenfold, entered tape-out in collaboration with TSMC.

Source: 2024-03-27 · reviewed 2026-09-24

Evidence: Primary-source statement · Reported operating

Limits: Establishes that tape-out began with TSMC as of March 2024. It does not establish the process node, volumes, pricing, that the chip reached volume production or flight, or that the relationship continues today; the source is more than two years old.

AST SpaceMobile ASIC Chip Enters Tape-Out Phase in Collaboration with TSMC ↗
TSMC manufactures chips for Amazon Web Services
Manufactures forCurrent
Industry analysts report that TSMC manufactures AWS's custom Trainium and Graviton chips on advanced process nodes, with ASIC design partner Alchip assisting; TSMC does not comment on individual customers.

Source: 2023-11-30 · reviewed 2026-09-22

Evidence: Secondary reporting (unconfirmed by primary parties) · Reported operating

Limits: Does not confirm exact process node, order volumes, or a direct contractual relationship, since TSMC declined to comment on individual customers; attribution rests on supply-chain analyst reporting.

Amazon Unveils New AWS-Designed Chips, Boosting Orders for TSMC and ALCHIP ↗
TSMC uses computational lithography technology from NVIDIA
Uses technology fromCurrent
TSMC integrated NVIDIA's cuLitho GPU-accelerated computational lithography software into its manufacturing software and process flow, as part of a joint NVIDIA/ASML/TSMC/Synopsys announcement, initially targeting Hopper-generation GPU production.

Source: 2023-03-21 · reviewed 2026-09-22

Evidence: Primary-source statement · Reported operating

Limits: Does not disclose how broadly cuLitho has been deployed across TSMC's fabs, financial terms, or whether coverage now extends beyond the initially announced Hopper-generation production.

NVIDIA, ASML, TSMC and Synopsys Set Foundation for Next-Generation Chip Manufacturing ↗
TSMC fabricates chips for Qualcomm
Manufactures forCurrent
Trade press has reported that TSMC manufactures Qualcomm's flagship Snapdragon mobile chips as a foundry partner, including sole-sourcing some recent Snapdragon generations.

Source: 2022-07-14 · reviewed 2026-09-20

Evidence: Secondary reporting (unconfirmed by primary parties) · Reported operating

Limits: This reflects industry reporting on foundry allocation for specific chip generations; it does not prove TSMC manufactures all current Qualcomm products, since Qualcomm has also used Samsung Foundry for some chips, and neither company has published exhaustive volume or exclusivity details.

TSMC to exclusively manufacture next-gen Qualcomm Snapdragon flagship chips ↗
TSMC certifies its process technology for design tools from Synopsys
Uses technology fromCurrent
Synopsys's digital and custom design platforms were certified for TSMC's N3 (3-nanometer) process technology, allowing chip designers to use Synopsys tools to design chips for that TSMC node.

Source: 2021-05-26 · reviewed 2026-09-22

Evidence: Primary-source statement · Reported operating

Limits: Certification of design tools for a process node does not indicate how many chips are actually taped out with these tools, deal value, or exclusivity versus other EDA vendors.

Synopsys Digital and Custom Design Platforms Certified for TSMC's Latest 3nm Process Technology ↗
TSMC manufactures data-infrastructure silicon for Marvell Technology
Manufactures forCurrent
TSMC and Marvell extended their partnership to deliver a broad silicon portfolio for the data infrastructure market on TSMC's N5P (5-nanometer) process technology, with multiple designs contracted and first products sampling by end of 2021.

Source: 2020-08-25 · reviewed 2026-09-22

Evidence: Primary-source statement · Reported operating

Limits: Describes a 2020 5nm-generation agreement; does not establish whether the same manufacturing arrangement covers Marvell's current 2nm custom-silicon designs.

Marvell and TSMC Collaborate to Deliver Industry's Most Advanced Data Infrastructure Portfolio on 5nm Technology ↗
TSMC certifies its process technology for design tools from Cadence Design Systems
Uses technology fromCurrent
Cadence's design solutions were certified for TSMC's A16 and N2P advanced process technologies, enabling AI and 3D-IC chip design on those nodes.

Source: Source undated · reviewed 2026-09-22

Evidence: Primary-source statement · Reported operating

Limits: Certification does not indicate specific customer tape-outs, revenue, or exclusivity between Cadence and TSMC; page content was accessed via search indexing rather than a direct fetch because cadence.com's robots rules blocked a live fetch in this session.

Cadence and TSMC Advance AI and 3D-IC Chip Design with Certified Design Solutions for TSMC's A16 and N2P Process Technologies ↗

Dated or announced (8)

TSMC is reported to fabricate MTIA chips for Meta Platforms
Manufactures forDated / announced
Reuters, citing an internal Meta memo, reported that TSMC manufactures Meta's next MTIA accelerator, which was slated to begin production in September 2026; Meta declined to comment.

Source: 2026-07-09 · reviewed 2026-09-24

Evidence: Secondary reporting (unconfirmed by primary parties) · Reported but unconfirmed

Limits: Based on an internal document rather than a statement by Meta or TSMC, and Meta declined to comment. It does not establish the process node, wafer volumes, contract value, or that production actually began on schedule.

Meta's new AI chips will begin production in September ↗
Amkor Technology provides advanced packaging and test manufacturing for TSMC
Manufactures forDated / announced
Amkor Technology and TSMC signed a 10-year partnership under which TSMC will procure advanced packaging and test services from Amkor, with both companies building facilities in Arizona to create a full U.S. semiconductor supply chain from silicon manufacturing to packaged devices.

Source: 2026-06-16 · reviewed 2026-09-22

Evidence: Primary-source statement · Announced / planned

Limits: Does not disclose contract value, packaging volumes, or the share of TSMC's total advanced-packaging demand Amkor will fulfill; the Arizona facilities are still under development.

TSMC and Amkor Technology Announce Long Term Partnership to Accelerate Advanced Packaging in the United States ↗
TSMC will fabricate AI5 chips for Tesla
Manufactures forDated / announced
Tesla completed tape-out of its AI5 chip in April 2026 and Elon Musk said the chip will be produced at both TSMC and Samsung, with volume manufacturing anticipated around 2027.

Source: 2026-04-16 · reviewed 2026-09-24

Evidence: Secondary reporting of an on-record executive statement · Announced; not confirmed operating

Limits: Tape-out is a design milestone, not production. This does not establish that TSMC has manufactured any AI5 volume, which TSMC fab or node will be used, the contract value, or how output will be split between TSMC and Samsung.

Tesla completes AI5 chip tape-out, to be manufactured by TSMC and Samsung ↗
ASML was recognized as a supplier by TSMC
SuppliesDated / announced
ASML received a 2024 TSMC supplier performance award.

Source: 2024-12-02 · reviewed 2026-09-17

Evidence: Primary-source statement · Supplier relationship documented

Limits: Supports a supplier relationship in 2024, not current contract terms, spend, exclusivity or dependency.

TSMC 2024 supplier awards ↗
Applied Materials was recognized as a supplier by TSMC
SuppliesDated / announced
Applied Materials received a 2024 TSMC supplier performance award.

Source: 2024-12-02 · reviewed 2026-09-17

Evidence: Primary-source statement · Supplier relationship documented

Limits: Supports a supplier relationship in 2024, not current contract terms, spend, exclusivity or dependency.

TSMC 2024 supplier awards ↗
KLA was recognized as a supplier by TSMC
SuppliesDated / announced
KLA received a 2024 TSMC supplier performance award.

Source: 2024-12-02 · reviewed 2026-09-17

Evidence: Primary-source statement · Supplier relationship documented

Limits: Supports a supplier relationship in 2024, not current contract terms, spend, exclusivity or dependency.

TSMC 2024 supplier awards ↗
Lam Research was recognized as a supplier by TSMC
SuppliesDated / announced
Lam Research received a 2024 TSMC supplier performance award.

Source: 2024-12-02 · reviewed 2026-09-17

Evidence: Primary-source statement · Supplier relationship documented

Limits: Supports a supplier relationship in 2024, not current contract terms, spend, exclusivity or dependency.

TSMC 2024 supplier awards ↗
Tesla dual-sources chip-foundry manufacturing from TSMC
Buys fromDated / announced
Reporting indicates Tesla is dual-sourcing fabrication of its AI5 chip across both TSMC and Samsung foundries for its 2026 global production ramp.

Source: Source undated · reviewed 2026-09-20

Evidence: Secondary reporting (unconfirmed by primary parties) · Reported dual-sourcing plan, 2026

Limits: The dual-sourcing claim has not been confirmed directly by Tesla or TSMC in an on-record statement; volumes and timing are not established by this source.

Tesla Breaks the Foundry Monopoly: Dual-Sourcing AI5 Silicon Across TSMC and Samsung's U.S. Fabs for 2026 Global Ramp ↗

Comparisons (1)

TSMC competes with Samsung Electronics
Competes withComparison only
TSMC and Samsung Electronics compete directly in advanced-node AI-accelerator foundry manufacturing; financial press analysis of Samsung's 2026 foundry push estimated TSMC holds roughly 95% of the AI-accelerator foundry market versus Samsung's 7-8%, with Samsung's 2nm yields (50-60%) still trailing TSMC's (80%+).

Source: 2026-07-30 · reviewed 2026-09-22

Evidence: Analytical comparison · Reported operating

Limits: Market-share and yield figures are this outlet's estimates, not disclosed by either company; does not establish exclusivity or predict the outcome of Samsung's competitive push.

Samsung's $200 Billion Broadcom Deal Is a Bid to Break TSMC's AI Foundry Lock ↗
Evidence for this profile

TSMC investor FAQ ↗
Reviewed

Explore it interactively

Open this profile in the database See it on the relationship map

Inclusion in this database is not an endorsement of a company or its securities. Nothing here is a recommendation to buy or sell anything, and none of it is individualized investment advice. See the disclosures. Data: AI Map v3.22, released .