Skip to content
GFS
~8 min read · 1,764 words ·updated 2026-04-29 · confidence 39%

Confidence legend: ✓ verified-primary (GF press release, SEC filing, 20-F Item 4.A) · ◐ partial / aggregator · ⚠ inferred / estimate.

GlobalFoundries’s M&A history is unusually concentrated in a small number of strategically defining transactions. Five corporate transactions account for the entire material record from formation through 2026-Q1: (1) the 2009 ATIC-led formation of GF as an AMD spin-out, (2) the 2010 Chartered Semiconductor Singapore acquisition, (3) the 2015 IBM Microelectronics acquisition (the unusual “we’ll pay you to take this” structure), (4) the 2022 Fab 10 East Fishkill divestiture to onsemi as part of the post-7nm-pivot specialty-foundry refocus, and (5) the 2025-11-17 Advanced Micro Foundry (AMF) Singapore acquisition that consolidates the merchant silicon-photonics foundry layer.

For the photonics-thesis scope on this KB site, the AMF acquisition is the load-bearing M&A event. Chartered (2010) provided the Singapore manufacturing base; AMF (2025) made GF the largest pure-play silicon-photonics foundry by revenue. Everything else is supporting context.


1. 2009 — ATIC Formation of GlobalFoundries (AMD Spin-Out)

  • Announced (parent transaction): 2008-10-07. AMD declares fabless transition; new entity to be created with backing from Mubadala-affiliated Advanced Technology Investment Company (ATIC).
  • Amended terms announced: 2008-12-08. AMD ~34.2% / ATIC ~65.8% post the $700M ATIC stake increase. ◐ Wikipedia — GlobalFoundries history
  • Officially launched: 2009-03-04 as GLOBALFOUNDRIES. Initial leadership: Doug Grose (CEO, former AMD SVP Manufacturing Operations); Hector Ruiz (Chairman, former AMD Executive Chairman). ✓ Mubadala press release — Worlds First Global Semiconductor Foundry Company Opens
  • Strategic rationale: AMD needed to monetize its capital-intensive manufacturing arm to focus on fabless design competition with Intel. ATIC / Mubadala saw the opportunity to build a pure-play foundry alternative to TSMC + UMC. The transaction created the first non-Asia-headquartered full-scale pure-play foundry.
  • Outcome: AMD’s residual ownership eventually fell to zero (~2012, exact date ⚠ confirm), leaving Mubadala-affiliated entities as 100% owner until the October 2021 IPO.

Key context for GFS investors today: This founding transaction is why GFS is structurally different from TSMC, Samsung Foundry, UMC, and SMIC. Each of those is a “born-foundry” entity with no prior IDM lineage. GFS inherited an IDM-era manufacturing footprint (multiple US fabs, broad process portfolio, x86-server-class capabilities) and has spent 17 years rationalizing that footprint into a coherent specialty-foundry strategy. The 2018 7nm halt was a pivotal step in that rationalization; the AMF acquisition is the most recent.


2. 2010 — Chartered Semiconductor Singapore Acquisition (~$3.9B)

  • Announced: 2009-09-07 by ATIC.
  • Closed / integration completed: 2010-01-13. ◐ GF Facebook — Chartered acquisition fact
  • Target: Chartered Semiconductor Manufacturing Co. Ltd. — Singapore-based, world’s #3 pure-play foundry at the time, headquartered in Woodlands Industrial Park, Kranji, Singapore. The Chartered site (now GF Singapore) was originally opened by Chartered in 1995 as one of the first 200mm fabs.
  • Deal structure: ~$3.9B total in cash + assumed debt (S$5.6B equivalent). Acquired by ATIC, then merged into GF as the Singapore manufacturing operation. ◐ Network World — ATIC to buy Chartered
  • Strategic rationale: Combine GF’s leading-edge US capabilities (Fab 8 ramp underway in Malta, NY) with Chartered’s mature-node specialty / Singapore supply-chain base. Achieve scale to compete more effectively with TSMC and UMC.
  • Outcome — long-term: The Singapore site became a workhorse for specialty / mature-node business after the 2018 7nm halt. It remains operationally significant today; the September 2023 expansion of the Singapore campus (a $4B GF capex program announced in 2021 and completed mid-2023) added incremental 200mm + 300mm capacity. ◐ CNBC — GF $4B Singapore expansion fab opens

Photonics-thesis-relevance: Pre-AMF (2025-11-17), the GF Singapore site was the legacy mature-node / specialty-CMOS Asia hub. Post-AMF, the GF Singapore presence is dual: legacy ex-Chartered + AMF silicon-photonics. The two operating units within Singapore are organizationally distinct (Yew Kong Tan SVP / GM APAC Manufacturing & Singapore Site has the legacy cluster; AMF integration is reporting under Kevin Soukup’s silicon-photonics business unit per inferred org structure). ⚠ confirm post-integration org chart in next refresh.


3. 2015 — IBM Microelectronics Acquisition (IBM Paid GF $1.5B)

  • Announced: 2014-10-20. ✓ GF press release — Acquisition Announced
  • Closed: 2015-07-01. ✓ GF press release — Acquisition Completed
  • Target: IBM’s commercial semiconductor technology business (“IBM Microelectronics”), including:
    • 200mm fab in Essex Junction, Vermont — became GF Fab 9.
    • 300mm fab in East Fishkill, New York — became GF Fab 10.
    • Intellectual property portfolio (~10,000+ patents).
    • World-class technologists / engineering teams.
  • Deal structure: IBM paid GlobalFoundries ~$1.5B over three years to take the business — a “negative consideration” structure rare in technology M&A. IBM took a $4.7B pre-tax charge in Q3 2014 financial results covering: asset impairment, estimated divestiture costs, and the $1.5B cash consideration to GF.
  • Supply agreement: GF became the sole supplier of IBM server-processor chips (z-series, POWER) for 10 years (until ~2025). ⚠ confirm post-2025 supply-arrangement renewal status in 20-F.
  • Strategic rationale:
    • For IBM: Exit unprofitable commercial-foundry business; concentrate R&D on systems / software / cloud strategy.
    • For GF: Acquire RF SOI capabilities (Burlington VT was the world-class RF SOI fab); access IBM’s vast patent portfolio (especially valuable for litigation-defense and licensing); pick up a 300mm logic fab; absorb world-class engineering talent.

Long-term outcome: The Burlington VT fab (Fab 9) became GF’s most important RF SOI specialty asset and is now central to the CHIPS Act expansion (the November 2024 CHIPS Act award supports a Vermont GaN-on-Si capacity expansion). The East Fishkill NY fab (Fab 10) was less strategically important post-7nm pivot and was divested in 2022 (see #4 below).

The negative-consideration structure of this acquisition is worth flagging for newcomer analysts: it means GF received cash for accepting a money-losing operation, not paid cash to acquire revenue. The sustainable economic value to GF therefore came from (a) the IP portfolio, (b) the Vermont RF SOI franchise, and (c) the long-term IBM supply commitment — not from immediate operating-income contribution.


4. 2022 — Fab 10 East Fishkill Divestiture to onsemi

  • Announced / closed: 2022 calendar year (exact dates ⚠ confirm in 20-F Item 4.A).
  • Buyer: ON Semiconductor Corp. (NASDAQ: ON; “onsemi”).
  • Asset: Fab 10 (East Fishkill, NY) — the 300mm logic fab inherited from IBM in 2015.
  • Strategic rationale:
    • For GF: Post-7nm-halt strategy emphasized differentiated specialty processes; East Fishkill’s general-purpose 300mm logic capacity didn’t align with the new playbook. Monetize the asset to fund Malta Fab 8 + Dresden expansions + Singapore expansion.
    • For onsemi: Acquire 300mm capacity for SiC and silicon power-semiconductor production aligned with EV / renewable-energy thesis.
  • Deal value: ⚠ confirm in 20-F + onsemi 10-K cross-reference.

Note on inheritance trail: Fab 10 was opened by IBM in the late 1990s, transferred to GF on 2015-07-01, transferred to onsemi in 2022. Three separate ownerships in ~24 years; a useful microcosm of the broader semiconductor industry’s capital-allocation dynamics around 300mm logic capacity.


5. 2025 — Advanced Micro Foundry (AMF) Singapore Acquisition

  • Announced / closed: 2025-11-17. ✓ GF press release — AMF Acquisition
  • Target: Advanced Micro Foundry Pte Ltd (“AMF”), Singapore — the world’s first specialty silicon-photonics pure-play foundry. AMF was originally a spin-out from Singapore’s A*STAR (Agency for Science, Technology and Research). 15+ years of silicon-photonics manufacturing expertise.
  • Deal structure: Financial terms NOT disclosed in the public press release. ⚠ confirm in next 20-F (FY2025) Item 4.A and 5.B for any disclosed purchase consideration / goodwill / intangibles.
  • Assets transferred:
    • Manufacturing facility: 200mm silicon-photonics platform in Singapore. Plans to scale to 300mm as market needs grow.
    • Intellectual property portfolio (in-house-developed Process Design Kits, evolving for Telecom / Data Center / LiDAR / Sensor applications).
    • Skilled engineering talent.
  • Strategic rationale:
    • For GF: Consolidate the merchant-photonics foundry layer. Becomes “the largest silicon photonics pure-play foundry by revenue.” Adds 200mm capacity (complementing Malta’s 300mm 45CLO Fotonix) for long-haul optical communications, computing, LiDAR, sensing applications. Establishes a Singapore-based Center of Excellence with A*STAR to develop next-generation materials targeting 400Gbps ultra-fast data transfer.
    • For AMF / Singapore ecosystem: Provides scale / global GTM reach for AMF’s process IP; aligns with Singapore’s national semiconductor-cluster strategy.
  • Concurrent transaction: GF also acquired InfiniLink in November 2025 — a separate compound-semiconductor / III-V integration capability (likely modulators, lasers, photodetectors). ⚠ details in GF blog post — Light-speed ambition ; deserves its own M&A entry once primary-source detail is extracted.

Photonics-thesis-relevance: This is the single most important M&A event in the GFS photonics narrative. It:

  1. Positions GF as the merchant-foundry leader for silicon photonics (vs Tower PH18, vs TSMC’s largely-captive SiPh, vs Intel-Jabil).
  2. Provides a “China-free” supply-chain narrative valuable to US / EU hyperscaler customers under the BIS export-control regime.
  3. Closes off AMF as an alternative foundry path for any photonics design house that previously had AMF as Plan B (notably affects POET Technologies, which historically has had multiple foundry partnerships outside the Sanan-IC / SPX China supply chain — POET’s foundry-allocation politics are directly affected by AMF being absorbed into GF).
  4. Adds 15+ years of merchant silicon-photonics process IP to GF’s portfolio, accelerating GF’s roadmap to 300mm advanced SiPh.

The “terms not disclosed” structure means analysts don’t have a price-paid number to evaluate ROIC. The next 20-F (FY2025, expected filing around 2026-03 next cycle, ⚠ date) is the first opportunity to triangulate the purchase consideration via goodwill, intangibles, and acquisition-related expense line items. Until then, AMF’s strategic value is qualitative and the financial impact is opaque.


What GF Has NOT Done

A discipline check on the bear-case argument that “GF is a serial acquirer”:

  • No M&A in 2016–2021 (six consecutive years between IBM close and IPO with zero acquisitions). The post-IBM era was integration / portfolio-rationalization, not platform-expansion.
  • No leveraged-buyout-style M&A. The 2010 Chartered transaction was equity-funded by ATIC; the 2015 IBM transaction was a “GF received cash” structure; the 2025 AMF transaction terms are undisclosed but presumably modest given AMF’s pre-acquisition revenue scale.
  • No US-domestic-foundry M&A (e.g., no GF-Tower combination, no GF-SkyWater roll-up). The merchant-US-foundry consolidation that some analysts have predicted has not happened on the GF side.

Cross-section pointers

  • ./timeline — Every M&A row above resolves to a dated timeline entry.
  • ./mubadala_control_structure — Mubadala / ATIC formation history; the founding transaction.
  • overview — Fotonix process detail; the AMF acquisition’s technical implications.
  • overview — The customer roster (LWLG, NLM, Marvell, Ayar Labs) that benefits from the AMF capacity addition.
  • overview — Capital-structure detail; the 2009 / 2010 / 2015 / 2025 transactions all have material balance-sheet implications still visible in the 20-F.