Mitsubishi Electric (4,602 patent records), Fuji Electric (4,551), and Toshiba (3,427) sit at the top of the IGBT module patent ranking, with Infineon Technologies AG (2,148) and Denso (1,896) rounding out a top five that holds 47% of the combined patent records of the hundred largest filers [S3].
That concentration defines the 2026 buying reality for engineers sourcing IGBT modules and [press-pack IGBT](https://www.hiitiosemi.com/de/blog/what-strategic-view-need-to-know-about-press-pack-igbt-in-2026/) devices: freedom-to-operate risk, long qualification cycles, and pin-compatible second-source pressure are now procurement variables, not background noise [S2][S3].
Patent concentration: who owns the IGBT IP core
The top three filers alone sit roughly 1,300 records ahead of the fourth-ranked Infineon Technologies AG (2,148), and the gap from rank five Denso (1,896) to rank six Rohm (1,610) is narrower than the gap from rank two to rank three, showing a pronounced top-heavy structure [S3]. Renesas Electronics (1,433), Sumitomo Electric Industries (1,349), Infineon Technologies Austria AG (990), and Toshiba Electronics Devices & Storage (981) fill the next tier, with Toyota Motor Corporation at 854 and Hitachi at 599 just behind [S3].
For a new entrant, this ranking translates into dense claim coverage on core IGBT device architectures; challengers are pushed toward adjacent branches such as wide-bandgap alternatives and advanced packaging, where incumbents hold thinner coverage [S3]. Annual filings peaked at 1,265 records in 2017 and eased to 974 by 2022, consistent with a maturing technology cycle rather than active growth, while 2024 and 2025 figures remain provisional because of publication lag [S3].
Press-pack IGBT: the 5.2–6.5 kV battleground
Press-pack IGBT procurement in 2026 centers on bond-wire-free, pressure-contact designs rated for 5.2–6.5 kV operation, with double-sided cooling and Short Circuit Failure Mode (SCFM) cited as the structural differentiators versus conventional wire-bond modules [S2]. SCFM provides a predictable short-circuit failure path instead of the catastrophic rupture pattern seen in bond-wire modules, a property that maps directly to HVDC, STATCOM, and VSC-HVDC converter uptime targets [S2].
Market entry in press-pack IGBT remains high-barrier because production is concentrated among legacy semiconductor manufacturers with decades of process know-how in pressure contact, hermetic housing, and spring clamping [S2]. For buyers, that concentration shapes three concrete 2026 decisions: pin-compatible second-source qualification, thermal cycling margin verification, and a longer-term hedge into SiC modules for new builds where the efficiency gap outweighs the reliability gap [S2].
Demand drivers: HVDC, offshore wind, AI data centers
2026 grid pressure is driven by renewable energy grid integration, HVDC transmission, and critical backup power, all of which demand higher voltage, higher reliability, and better thermal stability from power conversion stages [S2]. AI data center backup power and rail traction add load profiles that reward higher-voltage platforms and bond-wire-free failure modes, both of which are the headline selling points of press-pack IGBT [S2].
For spec-writing, this means a pressure transmitter on the cooling loop, a flow meter on the deionized water skid, and the PLC interlocking the converter cabinet are now selected alongside the IGBT, not after it, because the thermal envelope of a 6.5 kV press-pack module dictates the balance-of-plant design [S2]. The wider pattern, where SiC and GaN are reshaping the competitive map at lower voltages, is documented separately in a 2026 power-semiconductor roundup that complements the IGBT-specific patent data above.
Selection criteria: 5.2–6.5 kV press-pack vs. standard IGBT vs. SiC modules
The decision tree in 2026 breaks on three axes: voltage class, failure-mode tolerance, and second-source availability. A standard wire-bond IGBT module remains the cost-default for 1.7–3.3 kV industrial drives where footprint and short lead-time dominate; a 5.2–6.5 kV press-pack IGBT is the spec default for VSC-HVDC, STATCOM, and large offshore-wind converters where SCFM and thermal cycling margin are non-negotiable; a SiC module enters the comparison at 1.7–3.3 kV where switching frequency and efficiency outweigh absolute device cost [S2].
Procurement flexibility matters because the same converter cabinet is frequently re-bid across multiple vendors: press-pack IGBTs support pin-compatible OEM swaps that protect system teams from PCB re-spins when a single supplier exits a voltage class [S2]. The longer-horizon outlook for IGBT vs. SiC capex through 2030 is tracked in a separate demand-side analysis, and the balance-of-plant implications for cooling skids tie back to the same industrial valve selection gates used elsewhere on the converter platform [S2].
Standards, qualifications, and sourcing risk
Qualification cycles for new press-pack IGBT platforms are tight because grid and rail customers require multi-year thermal-cycling data, SCFM verification under short-circuit test, and HVDC-style dielectric validation at the target 5.2–6.5 kV class [S2]. Engineers specifying equipment in 2026 should treat 2024 and 2025 patent filing data as provisional and weight long-window concentration and technology-route coverage more heavily than the most recent annual bar [S3].
Two trackable signals to watch: first, whether any Tier-2 supplier breaks the top-five patent concentration by filing aggressively into wide-bandgap or advanced-packaging branches; second, whether a Tier-1 press-pack incumbent extends the 6.5 kV platform into a pin-compatible second-source family that lets buyers qualify two vendors without redesigning the stack [S2][S3].