Access Official Technical PDF

Share Summary:LinkedInEmail

The epitaxy equipment market is shaped by a broad shift in semiconductor manufacturing: silicon remains dominant, but applications with demanding performance requirements are creating opportunities for compound semiconductors. For automation engineers, system integrators and manufacturing leaders, that shift matters because different materials and devices call for different growth processes, equipment strategies and production conditions.

A 2020 market assessment from Yole Développement valued equipment for “More than Moore” epitaxy at nearly $940 million in 2019 and outlined an aggressive scenario in which the market would exceed $6 billion by 2025. These figures are historical estimates and a scenario, not a statement of present-day market size. The underlying drivers identified in the assessment—LEDs, power electronics and photonics—illustrate why forecasting equipment demand can be difficult.

Why epitaxy equipment demand is changing

Epitaxy grows a semiconductor layer on a substrate, enabling manufacturers to form material structures needed for particular devices. The process and equipment depend on the material system. Metal-organic chemical vapor deposition (MOCVD) is used across much of III–V compound semiconductor epitaxy, including gallium arsenide (GaAs) and gallium nitride (GaN) devices. High-temperature chemical vapor deposition (HT CVD) serves mainstream silicon-based components and silicon carbide (SiC) devices.

Silicon still accounts for more than 80% of the substrate market cited in the assessment. However, GaAs, GaN, SiC and indium phosphide (InP) are gaining relevance in applications where manufacturers seek performance that silicon solutions may not provide. Selecting a substrate is therefore an engineering and commercial decision: device requirements and material cost both matter.

Application requirements shape the process landscape

LEDs: established demand and new production challenges

GaN has been a major epitaxy market outside silicon, supported largely by conventional LED production. Yet the LED sector is not one uniform source of equipment demand. General lighting and backlighting had become commoditized, while manufacturers were exploring more specialized ultraviolet and infrared LEDs, as well as mini-LED and micro-LED displays.

Micro-LEDs raise the manufacturing bar. Their epitaxy requirements for defect control and material uniformity are more stringent than those for traditional LEDs. The assessment described equipment roadmaps aimed at roughly 0.1 defects per square centimeter or less for defects larger than one micrometer. It also noted tighter cleanroom operating conditions, including attention to automation and wafer cleaning. These demands are especially consequential for dies smaller than 10 micrometers, where smaller defects can become yield-limiting.

For production teams, this points to a distinction between simply adding capacity and meeting process-control requirements. Equipment investment for traditional GaN LEDs may not track wafer demand directly, while emerging display applications may depend on improvements in tool capability and manufacturing conditions.

SiC: power electronics and wide-bandgap materials

SiC substrates are attracting interest in power electronics, where reducing power consumption is important in areas such as transportation electrification, renewable energy, motor drives and some power supplies. Although SiC substrates are expensive, their properties make them a candidate for certain high-voltage metal-oxide-semiconductor field-effect transistor (MOSFET) and diode products.

The process landscape differs from that of GaN LED production: HT CVD was identified as the predominant approach for SiC devices. The assessment also noted efforts by MOCVD manufacturers to develop technologies for this market. For equipment planners, this is a reminder that growth in an application area does not automatically translate into demand for the same reactor technology across material systems.

VCSELs, laser diodes and RF applications

Photonics is another source of potential demand. Vertical-cavity surface-emitting lasers (VCSELs), commonly processed on GaAs, operate in the infrared spectrum and were making inroads into the epitaxy market. The consumer-goods industry’s adoption of VCSELs and edge-emitting laser diodes was also identified as a growth opportunity.

GaAs has a separate role in radio-frequency products. The assessment highlighted its use in small-cell applications for sub-6 GHz frequencies and early millimeter-wave deployments in the 28–39 GHz range. It anticipated that GaAs would remain important for sub-6 GHz 5G rather than CMOS silicon, citing requirements for power level and linearity alongside reduced antenna-board space, carrier aggregation and multiple-input multiple-output (MIMO) technology.

Equipment investment is not a simple demand forecast

Even where end-market demand appears promising, equipment orders can move differently from device shipments. In the period covered by the assessment, subsidies in China had contributed to substantial LED capacity, leaving MOCVD capacity for GaN LED production above actual production levels. The report cautioned that investment could be difficult to forecast and could vary from year to year, including if policy changes affected manufacturers’ production.

Traditional LED and micro-LED markets therefore present different planning questions. Conventional LED markets were described as increasingly commoditized, while micro-LED development depends on more demanding defect and homogeneity targets. Potential transitions from MOCVD to molecular beam epitaxy (MBE) could also affect equipment demand for compound semiconductor devices. The assessment pointed to possible MBE advantages in yield and uniformity for VCSELs and in high-frequency 5G RF applications; these were potential technology shifts, not settled outcomes.

Practical implications for manufacturing teams

For semiconductor operations and engineering leaders, the market picture suggests several considerations:

  • Match equipment to the material system. GaAs, GaN, SiC and silicon applications do not share one universal epitaxy process.
  • Assess process requirements alongside capacity. Micro-LED defect and homogeneity targets bring cleanroom, wafer-cleaning and automation considerations into equipment planning.
  • Separate established demand from emerging applications. Traditional LEDs, power devices, VCSELs and micro-LEDs have distinct manufacturing needs and investment uncertainties.
  • Track technology transitions. Possible movement between MOCVD, HT CVD and MBE can influence equipment choices, but should be evaluated against the specific device and process requirements.

The practical takeaway is that the epitaxy equipment market cannot be understood through a single growth forecast. Demand depends on which substrate and device applications gain traction, the production capabilities they require, and how manufacturers respond to existing capacity. For plant and engineering teams, investment decisions are strongest when they connect application outlooks to material choice, process control and the actual requirements of the intended device.

Access Official Technical PDF

About The Author