BuildNot yet confirmed elsewhere1 publisher2 min readPublished
GlobalFoundries plans 7nm-class planar FD-SOI production in Dresden in 2028
GlobalFoundries unveiled FDX Fusion, a 7nm-class planar FD-SOI node due in production at Dresden in 2028. It claims over twice 22FDX's transistor density for automotive, edge AI and industrial chips that value low power and cost over peak performance.
The Engineer · Build desk
Four figures. From PDK to FDX Fusion production start: at most roughly 18 months. FDX Fusion RF transistors: fT/fmax up to 0.5 THz. From the 12FDX announcement in 2016 to FDX Fusion production start: 12 years. 12FDX promised up to 50% lower power.
| Figure | Value | Note | Claim |
|---|---|---|---|
| From PDK to production start, at most roughly | 18 months | 18 | |
| RF fT/fmax, up to | 0.5 THz | 7 | |
| From 12FDX announcement to FDX Fusion production start | 12 years | 12FDX announced in 2016 | 17 |
| Promised by 12FDX, up to | 50 % lower power | 12 |
What happened
- GF plans demo silicon for customer evaluation in early 2027 and the first process design kit in mid-2027.
- The node is meant to put RF transistors reaching 0.5 THz fT/fmax, analog circuitry, embedded nonvolatile memory and in-memory computing on one die.
- GF's release lists BMW, Bosch, Infineon, NXP, Siemens and Tesla as interested, but only Infineon confirmed collaboration, on radar and wireless.
- GF announced 12FDX in 2016, promising 10nm FinFET-class performance and up to 50% lower power, with first customer tape-outs expected in 2019.
Why it matters
- decision A team targeting 2028 production gets at most about 18 months from PDK to production start, so the choice to wait for FDX Fusion or stay on a current process falls on early-2027 demo silicon.
- exposure Early adopters carry the schedule risk, because GF has not identified products or committed volumes with the customers it is working with.
- contradiction 12FDX stayed on GF's 2022 investor roadmap although Tom's Hardware says it was never adopted for a commercial product, so a roadmap entry is weak evidence that a node will ship.
The label is the least checkable part of the announcement. GF's one quantitative digital claim is relative: more than twice the transistor density of 22FDX, which dates from circa 2015 [3]. GF has not disclosed transistor dimensions, contacted gate pitch, minimum metal pitch or actual density [4]. Tom's Hardware calls GF's 7nm-class digital performance claim ambiguous [15]. For the label to transfer to a given chip, the pitches would have to land near those of FinFET nodes, and that chip's logic would have to scale with them. Until the pitches are published, a designer cannot test either condition.
The integration case is easier to evaluate. Tom's Hardware argues that the main selling point is putting relatively powerful digital logic on one die with RF, analog and mixed-signal circuitry, embedded nonvolatile memory and in-memory computing. Those functions would otherwise need several chips on different processes, with extra tapeout cost and power [9]. In our view the case weakens for any block that would have done better on its own process, because the combined die has to accept the compromise. GF plans processor IP based on MIPS, ARC and RISC-V to speed development [8].
GF dropped leading-edge development in 2018, canceling a 7nm-class FinFET process it co-developed with IBM [1]. The gap between the 12FDX announcement and FDX Fusion's planned production start is 12 years [17].
Sébastien Dauvé, chief executive of CEA Leti, welcomed the new node. "Through the FAMES pilot line, CEA-Leti continues to shape the future European roadmap for FD-SOI and works every day to promote its adoption in a wide range of applications," he said [14].
What to watch
- Publication of contacted gate pitch, minimum metal pitch and transistor density, which would let outsiders place the node against FinFET processes.
- Whether early-2027 demo silicon and the mid-2027 PDK arrive on schedule.
- Whether any customer beyond Infineon confirms a product or committed volume on FDX Fusion.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence35
- Adoption8
- Hype gap+40
- Incentives
- Insufficient
- Confidence60
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
When GlobalFoundries abandoned leading-edge node development in 2018, it also canceled its 7nm-class FinFET process co-developed with IBM.
- [2]
FDX Fusion is a 7nm-class process built on planar transistors and FD-SOI, aimed at automotive, edge AI, industrial, robotics and sensor devices that prioritize low power, high integration and cost efficiency over maximum performance.
- [3]
FDX Fusion promises 7nm-class digital performance and more than twice the transistor density of the original 22FDX process, introduced circa 2015.
- [4]
GF has not disclosed transistor dimensions, contacted gate pitch, minimum metal pitch or actual transistor density for FDX Fusion.
- [5]
GF plans to launch production of chips on FDX Fusion at its Dresden, Germany, facility in 2028.
- [6]
GF plans initial demo silicon for customer evaluation in early 2027, followed by the first process design kit (PDK) in mid-2027.
- [7]
FDX Fusion will support RF transistors with fT/fmax of up to 0.5 THz, analog and mixed-signal circuitry, embedded nonvolatile memory and in-memory computing.
- [8]
GF intends to provide processor IP based on MIPS, ARC and RISC-V architectures to accelerate development of FDX Fusion products.
- [9]
Tom's Hardware argues the main selling point of FDX Fusion is combining relatively powerful digital logic with RF, analog and mixed-signal circuitry, embedded nonvolatile memory and in-memory computing on one chip, which would otherwise require multiple chips on different process technologies, meaning additional tapeout costs and power consumption.
- [10]
GF is collaborating with several customers on FDX Fusion but has not identified specific products or committed production volumes.
- [11]
GF's press release said BMW, Bosch, Infineon, NXP, Siemens and Tesla have expressed interest in FDX Fusion, but only Infineon specifically confirmed collaboration on future 7nm-class technologies for radar and wireless communications.
- [12]
GF announced 12FDX in 2016, promising performance comparable to 10nm FinFET, full-node density improvements and up to 50% lower power consumption, with initial customer tape-outs expected in 2019.
- [14]
"Through the FAMES pilot line, CEA-Leti continues to shape the future European roadmap for FD-SOI and works every day to promote its adoption in a wide range of applications," said Sébastien Dauvé, chief executive of CEA Leti.
- [15]
Tom's Hardware describes GF's 7nm-class digital performance claim as ambiguous.
- [16]
12FDX has never been adopted for commercial products.
- [17]
FDX Fusion's planned production start (2028) comes 12 years after the 12FDX announcement (2016).
- [18]
From the mid-2027 PDK to a 2028 production start is at most roughly 18 months.
Sources
1 independent publisher whose own reporting we read for this story.
- tomshardware.comGlobalFoundries unveils 7nm-class FDX Fusion node — comes 7 years after firm abandoned FinFET
1 article · October 12, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.
Topics
- European chip manufacturingFollow
- Automotive SemiconductorsFollow
- FD-SOI process technologyFollow
- Semiconductor foundriesFollow