Semiconductors
SiC vs. GaN: why the application requirements should lead the decisions
Wide-bandgap value is not realized by material choice alone. It comes from matching device technology, package, and implementation within the system’s real operating demands. By Filippo Scrimizzi, system applications director, Nexperia, and Jim Honea, system applications engineering director, Nexperia
C
omparison is one of the ways we make sense of a complex decision. For engineers, that instinct is reinforced by the need to benchmark options, isolate
trade-offs, and determine which performs better against a defined requirement. It is therefore understandable that the power semiconductor conversation can often begin with a deceptively simple question: Silicon carbide (SiC) or Gallium nitride (GaN)? The comparison is useful, but only up to a point. A device is not selected to win a contest between materials. It is selected to meet a particular combination of electrical, thermal, mechanical, and lifecycle requirements within a real power system. That distinction matters because the material creates capability, not a guaranteed system outcome. Higher switching speed, lower switching losses, or stronger high- voltage performance can all be valuable, but only when they support the application’s operating envelope. If the voltage rating, load profile, thermal path, layout, package, or qualification strategy does not work, a theoretically attractive device can become completely the wrong system choice. The better starting point is therefore not “Which material is best?” It is “What must the application achieve, under which conditions, and for how long?” That question changes SiC vs. GaN from a contest into an engineering framework.
The operating envelope narrows
An application-first assessment defines the constraints before comparing device families. Blocking voltage and transient headroom are important engineering starting points, but they represent only one layer of the decision.
30 October 2026
requires 1200 V devices or a multilevel approach. The operating envelope does not always produce a single, obvious answer. What it does is remove false certainty and make the trade-offs visible. That is a more useful basis for wide-bandgap (WBG) device selection than assigning a fixed territory to each material.
Once the application requirements are understood, the next step is not simply to assign SiC to high voltage and GaN to high frequency. Those tendencies are useful, but they do not explain how the devices behave in a circuit.
The answer depends first on the electrical demands of the application, including its power level, switching behaviour, topology, and load profile. Thermal and mechanical constraints then shape what is practical, from ambient and junction temperature limits to the available cooling, physical size, and target power density. Beyond immediate performance, expected lifetime, qualification requirements, and manufacturing strategy determine whether a device is suitable for the product as a whole.
The balance between these requirements changes from one application to the next. A compact server power supply may benefit from high-frequency operation if it reduces the size of the magnetics and frees valuable board area. A high-voltage electric vehicle (EV) charging stage may place greater weight on voltage margin, sustained-load efficiency, and predictable behaviour at elevated temperatures. An industrial motor drive may prioritize ruggedness, long service life, and thermal performance across
Components in Electronics
www.cieonline.co.uk
variable loads. Each is a power-conversion application, but each asks something different of the semiconductor. Even within one application category, topology and mission profile can shift the balance. A 650 V device class may be relevant in one architecture, while another
SiC: voltage capability is only the first filter
SiC is particularly relevant in high-power conversion stages that must operate efficiently under sustained load. 650 to 1200 V SiC MOSFETs can address
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48
