Fuji Electric Unveils Next-Generation IGBT

Fuji Electric Announces New Generation IGBT – Adding a "Si Evolution" Option to the SiC-Dominant Market

In the mid-2020s, when silicon (Si) IGBTs (Insulated Gate Bipolar Transistors) are increasingly deemed obsolete, Fuji Electric has unveiled a new generation of IGBTs. Positioned as the company's "8th generation," these devices reportedly achieve significant improvements in both switching loss and on-voltage compared to previous generations. Fuji Electric is poised to directly compete in zones that overlap with SiC, such as EV inverters and industrial power supplies.

The question "Why IGBTs now, not SiC?" extends beyond a simple choice of materials. The answer is far more complex than a mere performance comparison of individual devices, encompassing cost structures, supply stability, and system design constraints.

Why IGBTs Can Still Compete

Amidst the rapid expansion of SiC power semiconductor adoption, IGBTs continue to maintain a significant market share due to cost and manufacturing infrastructure. SiC substrates are several times, or even more than ten times, more expensive than Si wafers, and their mass production systems are still developing. The industry as a whole is in a race toward 200mm wafer transition, as exemplified by Mitsubishi Electric's joint development of 8-inch SiC substrates with Coherent. The realistic view on the manufacturing floor is that the competitiveness of Si IGBTs will not diminish until this transition is complete.

Fuji Electric's 8th generation IGBTs are targeting precisely this "transitional gap." For the next two to five years, until SiC becomes truly cost-competitive, improved Si IGBTs can serve as a strong option for system designers. As Toshiba Device & Storage demonstrated in 2021 by reducing losses by up to 40.5% with their triple-gate IGBT, Si technology itself continues to evolve. Fuji Electric's move appears to be in line with this trend, although details regarding the structural improvements adopted in the 8th generation require further scrutiny of technical documentation at present.

Breakdown of Loss Reduction and "Where the Difference Lies"

The core metrics for evaluating IGBT performance are switching loss (Esw) and on-voltage (VCEsat). These two are fundamentally in a trade-off relationship. Reducing the on-voltage decreases conduction loss but increases switching loss. The priority between the two depends on the application's operating frequency and current profile.

This structure presents essentially the same problem as the trade-off between short-circuit withstand time and on-resistance in SiC MOSFETs.

For SiC, technologies that mitigate this trade-off through device structure improvements have emerged from various manufacturers. For IGBTs, similar approaches, namely precise control of carrier profiles and optimization of gate structures, are believed to be the focus of improvements in the 8th generation. While the specific structural technologies Fuji Electric employs require confirmation through datasheets and technical papers, the industry's common challenge is to "shift the boundaries" of this trade-off.

Where the Segmentation with SiC Will Be Determined

The question of "SiC or IGBT" is often discussed in the context of performance comparison. However, in actual design and procurement scenarios, the decision-making process is more complex, involving factors such as operating frequency, power supply voltage levels, production volume, and supplier supply stability.

In terms of voltage levels, current SiC devices are seeing their portfolios established in the 650V to 1700V range. For instance, onsemi offers SiC MOSFETs, SiC diodes, and SiC modules from 650V to 1700V. If Fuji Electric's 8th generation IGBTs primarily target the same band, differentiation based on voltage range may be difficult. Instead, system temperature conditions and protection circuit design costs are likely to become deciding factors.

A physical challenge with SiC is that the dies are smaller and current density is higher, leading to faster temperature rises compared to Si devices.

This characteristic increases the design burden for protection circuits and also impacts gate driver selection. Si IGBTs offer greater design margins in this regard and allow for easier reuse of existing protection circuit assets. The possibility of transitioning to new generation IGBTs as an "extension of existing designs" rather than a complete redesign is a significant consideration, particularly for projects prioritizing mass production transition costs.

Fuji Electric's Position and Competitors' Moves

In Japan's power semiconductor market, Fuji Electric, Mitsubishi Electric, and Toshiba stand as major players in the IGBT sector. While each company continues to invest in SiC technology—Mitsubishi Electric, for example, improved short-circuit withstand capability by introducing a p-type protective layer in its SiC trench structure—they are also consistently updating their Si IGBT generations.

The SiC camp has also refined its value propositions by focusing on trade-off improvements, as demonstrated by Rohm's claim of achieving both low on-resistance and high short-circuit withstand capability with its 4th generation SiC MOSFETs. Amidst this competitive landscape, Fuji Electric's timing in launching new generation IGBTs aligns with a strategic intent to "solidify market share before SiC's dominance is established."

This graph illustrates the natural segmentation between IGBTs and SiC based on operating frequency. The improvements offered by new generation IGBTs are most effective in the under-20kHz range, while SiC's lower switching loss tends to provide an advantage at higher frequencies. Organizing operating conditions by application facilitates decision-making on which technology is more suitable.

What to Consider Now in Both Technology Selection and Procurement

The emergence of new generation IGBTs will not immediately overturn existing SiC transition plans, but it may prompt a re-evaluation of options. For development projects anticipating product launches between 2026 and 2028, if the device selection has so far been solely focused on SiC, there is room for renewed comparative consideration.

Organizing the decision-making criteria from both technical and procurement perspectives reveals several layers.

As details emerge regarding which applications Fuji Electric will prioritize for its 8th generation IGBTs, and how it will set mass production timing and pricing, the landscape of SiC versus IGBT may gradually be redrawn. Rather than a binary "who will win" dichotomy, the practical decision-making criterion in this phase will be to refine the question of "under which conditions should each be used."