The silicon carbide industry is sending mixed signals.
In July 2026, a semiconductor facility in the United States began sample production of silicon carbide power chips. Only a few days later, a separate U.S. operation announced that it would shut down its silicon carbide wafer business, affecting about 140 employees.
At the same time, the technology is moving in the opposite direction. The industry has reported new progress in 300 mm, or 12-inch, single-crystal silicon carbide wafers, along with equipment designed for 12-inch epitaxy, thinning and substrate processing.
So which story is correct?
Is silicon carbide entering a downturn, or is it preparing for its next stage of growth?
The short answer is that both can be true.
Silicon carbide is not disappearing. It is moving from a shortage-driven market into a more demanding manufacturing cycle—one defined by yield, cost, customer qualification and the ability to scale.
A silicon carbide wafer is not the same thing as a silicon carbide chip.
The complete manufacturing chain includes several distinct steps, and each has its own technical and economic challenges.
High-purity silicon carbide source material is heated to temperatures above 2,000°C. The material sublimates and then recrystallizes into a single-crystal boule.
This process is much slower and more difficult to control than conventional silicon crystal growth. Micropipes, dislocations, polytype inclusions and internal stress can reduce the amount of usable material.
Larger crystal diameters make temperature distribution and stress control even more challenging.
The boule is oriented, sliced, ground, thinned and polished into individual substrates.
Thickness variation, bow, warp, scratches, subsurface damage and crystal defects must all be carefully controlled. Problems introduced at this stage can reduce yields throughout the rest of the production process.
A carefully controlled silicon carbide epitaxial layer is grown on the substrate.
This layer contains the electrical structure needed for the future device. Its thickness, doping uniformity and defect density help determine the voltage rating, efficiency and reliability of the finished component.
The epitaxial wafer then moves through ion implantation, lithography, etching, oxidation, metallization and electrical testing.
After fabrication, the wafer is diced into individual devices and packaged as MOSFETs, diodes or power modules.
A company producing substrates therefore operates in a different market from a company manufacturing power devices. They may serve different customers, use different equipment and face different cost pressures.
That is why the closure of a substrate operation does not mean that demand for every silicon carbide device has disappeared. Likewise, the launch of a new device production line does not prove that the entire market has returned to a shortage.
Several years ago, the outlook for silicon carbide appeared straightforward.
Electric vehicle sales were rising, 800-volt vehicle platforms were gaining attention, and silicon carbide was expected to become increasingly common in traction inverters, onboard chargers and high-voltage DC-DC converters.
Manufacturers across the supply chain responded by building new crystal-growth, substrate, epitaxy and device capacity.
But semiconductor capacity takes years to bring online.
By the time some of these projects were ready, electric vehicle growth in several markets had slowed compared with earlier forecasts. Automakers were also placing greater pressure on suppliers to reduce component costs.
The result was a mismatch: production capacity expanded faster than near-term demand.
This was particularly important for 150 mm, or 6-inch, silicon carbide substrates. As more capacity entered the market and manufacturing knowledge became more widely available, a product that had once been scarce began facing much stronger price competition.
The physics of silicon carbide did not suddenly stop working. The commercial assumptions simply changed.
The market is moving from a question of “Who can make SiC?” to a harder question:
Who can make it consistently, at high yield and at a price customers will accept?
Silicon carbide remains valuable because it can perform efficiently in conditions that are difficult for conventional silicon power devices.
Compared with silicon, silicon carbide offers a wider bandgap, higher critical electric field and better thermal conductivity. These properties can support higher voltages, faster switching and operation at elevated temperatures.
At the system level, properly designed silicon carbide electronics can reduce conversion losses and allow engineers to use smaller magnetic components, capacitors and cooling systems.
That makes the material attractive for applications including:
However, silicon carbide is not automatically the best choice for every product.
For lower-voltage or cost-sensitive equipment, mature silicon devices may deliver adequate performance at a lower price. Silicon carbide creates the greatest value when efficiency, voltage, switching speed, thermal performance or system size justifies the additional cost.
The next phase of the market will depend less on promoting silicon carbide as a universal replacement and more on identifying the applications where it produces a measurable system-level benefit.
Wafer size has become one of the most closely watched parts of the silicon carbide cost roadmap.
| Wafer size | Diameter | Current position | Main challenge |
|---|---|---|---|
| 6-inch | 150 mm | Established volume production | Price competition and cost reduction |
| 8-inch | 200 mm | Early commercial ramp and qualification | Yield, defect control and process stability |
| 12-inch | 300 mm | Demonstration and development stage | Crystal quality, equipment readiness and economics |
Moving from 6-inch to 8-inch increases the theoretical wafer area by about 78%. If yields remain stable, the larger wafer can produce more devices per manufacturing cycle and may reduce the cost allocated to each chip.
But larger wafers do not automatically create cheaper chips.
An 8-inch crystal is more difficult to grow uniformly. Manufacturers must also control wafer thickness, bow, warp, surface damage and epitaxial uniformity across a larger area.
If defect density increases or device yield falls, the additional wafer area may not translate into lower cost.
That is why the success of 8-inch silicon carbide will be measured by repeatable high-volume production—not by the number of sample wafers displayed at industry events.
In early 2026, the industry reported a new milestone involving a single-crystal 300 mm silicon carbide wafer.
Other recent developments have included 12-inch conductive and semi-insulating substrate samples, 12-inch epitaxial processing equipment and new systems for thinning large silicon carbide boules and substrates.
These developments are important because they show that 12-inch silicon carbide is moving beyond a theoretical proposal. Different parts of the supply chain are beginning to prepare for the format.
However, producing a sample wafer is very different from producing thousands of qualified wafers every month.
A commercial 12-inch platform would need to demonstrate:
For now, the most accurate description is that 6-inch silicon carbide is in mature production, 8-inch is entering a commercial ramp, and 12-inch remains primarily in technology development and process validation.
A 12-inch wafer has approximately four times the theoretical area of a 6-inch wafer and 2.25 times the area of an 8-inch wafer.
In principle, this allows many more devices to be processed in a single production cycle. It could also align silicon carbide with the 300 mm equipment infrastructure already widely used in the conventional silicon industry.
But silicon carbide is not simply another type of silicon wafer.
It is harder, more brittle and more difficult to grow. Large-diameter crystals can develop thermal stress, structural defects and geometry problems. Cutting and polishing the material is also time-consuming and can waste expensive crystal material.
As a result, the economics of 12-inch silicon carbide will depend on yield rather than diameter alone.
A large wafer with poor crystal quality or frequent breakage may cost more per usable device than a smaller wafer produced on a stable, mature line.
Another important part of the recent 12-inch activity is that not every program is focused exclusively on conventional power semiconductors.
Artificial intelligence and high-performance computing systems are creating new challenges in thermal management, package size and power density.
Because silicon carbide combines strong thermal performance with mechanical stability and useful electrical properties, large-format SiC is also being evaluated for advanced packaging, thermal management, photonic integration and other high-power computing applications.
This creates two possible development paths.
The first is the traditional power semiconductor path, where larger wafers could eventually reduce the cost of MOSFETs, diodes and other high-voltage devices.
The second is a broader materials platform for AI infrastructure, high-performance computing, photonics and advanced packaging.
These newer applications are still largely in the evaluation stage. They should not yet be treated as established volume markets. But they could expand the long-term role of silicon carbide beyond electric vehicles.
The most important silicon carbide indicators will no longer be announced capacity or wafer diameter alone.
Instead, customers and investors will need to watch:
These factors will determine whether larger wafers become a commercial advantage or simply a technical demonstration.
The silicon carbide industry is not collapsing, but its business model is changing.
Some capacity built during the shortage period may be consolidated or shut down. At the same time, the technical roadmap continues to move from 6-inch to 8-inch wafers, with early exploration of 12-inch platforms.
These trends are not contradictory.
They show that silicon carbide is leaving the phase in which almost any new capacity could attract attention. The industry is entering a more disciplined manufacturing cycle in which yield, quality, cost and real customer demand matter more than expansion announcements.
Six-inch production will continue to shape today’s market. Eight-inch manufacturing will influence the next stage of cost competition. Twelve-inch technology represents a longer-term frontier that could eventually reshape both power electronics and advanced computing.
Silicon carbide is not leaving the stage. It is simply entering the part of the technology cycle where manufacturing execution matters more than market excitement.
The silicon carbide industry is sending mixed signals.
In July 2026, a semiconductor facility in the United States began sample production of silicon carbide power chips. Only a few days later, a separate U.S. operation announced that it would shut down its silicon carbide wafer business, affecting about 140 employees.
At the same time, the technology is moving in the opposite direction. The industry has reported new progress in 300 mm, or 12-inch, single-crystal silicon carbide wafers, along with equipment designed for 12-inch epitaxy, thinning and substrate processing.
So which story is correct?
Is silicon carbide entering a downturn, or is it preparing for its next stage of growth?
The short answer is that both can be true.
Silicon carbide is not disappearing. It is moving from a shortage-driven market into a more demanding manufacturing cycle—one defined by yield, cost, customer qualification and the ability to scale.
A silicon carbide wafer is not the same thing as a silicon carbide chip.
The complete manufacturing chain includes several distinct steps, and each has its own technical and economic challenges.
High-purity silicon carbide source material is heated to temperatures above 2,000°C. The material sublimates and then recrystallizes into a single-crystal boule.
This process is much slower and more difficult to control than conventional silicon crystal growth. Micropipes, dislocations, polytype inclusions and internal stress can reduce the amount of usable material.
Larger crystal diameters make temperature distribution and stress control even more challenging.
The boule is oriented, sliced, ground, thinned and polished into individual substrates.
Thickness variation, bow, warp, scratches, subsurface damage and crystal defects must all be carefully controlled. Problems introduced at this stage can reduce yields throughout the rest of the production process.
A carefully controlled silicon carbide epitaxial layer is grown on the substrate.
This layer contains the electrical structure needed for the future device. Its thickness, doping uniformity and defect density help determine the voltage rating, efficiency and reliability of the finished component.
The epitaxial wafer then moves through ion implantation, lithography, etching, oxidation, metallization and electrical testing.
After fabrication, the wafer is diced into individual devices and packaged as MOSFETs, diodes or power modules.
A company producing substrates therefore operates in a different market from a company manufacturing power devices. They may serve different customers, use different equipment and face different cost pressures.
That is why the closure of a substrate operation does not mean that demand for every silicon carbide device has disappeared. Likewise, the launch of a new device production line does not prove that the entire market has returned to a shortage.
Several years ago, the outlook for silicon carbide appeared straightforward.
Electric vehicle sales were rising, 800-volt vehicle platforms were gaining attention, and silicon carbide was expected to become increasingly common in traction inverters, onboard chargers and high-voltage DC-DC converters.
Manufacturers across the supply chain responded by building new crystal-growth, substrate, epitaxy and device capacity.
But semiconductor capacity takes years to bring online.
By the time some of these projects were ready, electric vehicle growth in several markets had slowed compared with earlier forecasts. Automakers were also placing greater pressure on suppliers to reduce component costs.
The result was a mismatch: production capacity expanded faster than near-term demand.
This was particularly important for 150 mm, or 6-inch, silicon carbide substrates. As more capacity entered the market and manufacturing knowledge became more widely available, a product that had once been scarce began facing much stronger price competition.
The physics of silicon carbide did not suddenly stop working. The commercial assumptions simply changed.
The market is moving from a question of “Who can make SiC?” to a harder question:
Who can make it consistently, at high yield and at a price customers will accept?
Silicon carbide remains valuable because it can perform efficiently in conditions that are difficult for conventional silicon power devices.
Compared with silicon, silicon carbide offers a wider bandgap, higher critical electric field and better thermal conductivity. These properties can support higher voltages, faster switching and operation at elevated temperatures.
At the system level, properly designed silicon carbide electronics can reduce conversion losses and allow engineers to use smaller magnetic components, capacitors and cooling systems.
That makes the material attractive for applications including:
However, silicon carbide is not automatically the best choice for every product.
For lower-voltage or cost-sensitive equipment, mature silicon devices may deliver adequate performance at a lower price. Silicon carbide creates the greatest value when efficiency, voltage, switching speed, thermal performance or system size justifies the additional cost.
The next phase of the market will depend less on promoting silicon carbide as a universal replacement and more on identifying the applications where it produces a measurable system-level benefit.
Wafer size has become one of the most closely watched parts of the silicon carbide cost roadmap.
| Wafer size | Diameter | Current position | Main challenge |
|---|---|---|---|
| 6-inch | 150 mm | Established volume production | Price competition and cost reduction |
| 8-inch | 200 mm | Early commercial ramp and qualification | Yield, defect control and process stability |
| 12-inch | 300 mm | Demonstration and development stage | Crystal quality, equipment readiness and economics |
Moving from 6-inch to 8-inch increases the theoretical wafer area by about 78%. If yields remain stable, the larger wafer can produce more devices per manufacturing cycle and may reduce the cost allocated to each chip.
But larger wafers do not automatically create cheaper chips.
An 8-inch crystal is more difficult to grow uniformly. Manufacturers must also control wafer thickness, bow, warp, surface damage and epitaxial uniformity across a larger area.
If defect density increases or device yield falls, the additional wafer area may not translate into lower cost.
That is why the success of 8-inch silicon carbide will be measured by repeatable high-volume production—not by the number of sample wafers displayed at industry events.
In early 2026, the industry reported a new milestone involving a single-crystal 300 mm silicon carbide wafer.
Other recent developments have included 12-inch conductive and semi-insulating substrate samples, 12-inch epitaxial processing equipment and new systems for thinning large silicon carbide boules and substrates.
These developments are important because they show that 12-inch silicon carbide is moving beyond a theoretical proposal. Different parts of the supply chain are beginning to prepare for the format.
However, producing a sample wafer is very different from producing thousands of qualified wafers every month.
A commercial 12-inch platform would need to demonstrate:
For now, the most accurate description is that 6-inch silicon carbide is in mature production, 8-inch is entering a commercial ramp, and 12-inch remains primarily in technology development and process validation.
A 12-inch wafer has approximately four times the theoretical area of a 6-inch wafer and 2.25 times the area of an 8-inch wafer.
In principle, this allows many more devices to be processed in a single production cycle. It could also align silicon carbide with the 300 mm equipment infrastructure already widely used in the conventional silicon industry.
But silicon carbide is not simply another type of silicon wafer.
It is harder, more brittle and more difficult to grow. Large-diameter crystals can develop thermal stress, structural defects and geometry problems. Cutting and polishing the material is also time-consuming and can waste expensive crystal material.
As a result, the economics of 12-inch silicon carbide will depend on yield rather than diameter alone.
A large wafer with poor crystal quality or frequent breakage may cost more per usable device than a smaller wafer produced on a stable, mature line.
Another important part of the recent 12-inch activity is that not every program is focused exclusively on conventional power semiconductors.
Artificial intelligence and high-performance computing systems are creating new challenges in thermal management, package size and power density.
Because silicon carbide combines strong thermal performance with mechanical stability and useful electrical properties, large-format SiC is also being evaluated for advanced packaging, thermal management, photonic integration and other high-power computing applications.
This creates two possible development paths.
The first is the traditional power semiconductor path, where larger wafers could eventually reduce the cost of MOSFETs, diodes and other high-voltage devices.
The second is a broader materials platform for AI infrastructure, high-performance computing, photonics and advanced packaging.
These newer applications are still largely in the evaluation stage. They should not yet be treated as established volume markets. But they could expand the long-term role of silicon carbide beyond electric vehicles.
The most important silicon carbide indicators will no longer be announced capacity or wafer diameter alone.
Instead, customers and investors will need to watch:
These factors will determine whether larger wafers become a commercial advantage or simply a technical demonstration.
The silicon carbide industry is not collapsing, but its business model is changing.
Some capacity built during the shortage period may be consolidated or shut down. At the same time, the technical roadmap continues to move from 6-inch to 8-inch wafers, with early exploration of 12-inch platforms.
These trends are not contradictory.
They show that silicon carbide is leaving the phase in which almost any new capacity could attract attention. The industry is entering a more disciplined manufacturing cycle in which yield, quality, cost and real customer demand matter more than expansion announcements.
Six-inch production will continue to shape today’s market. Eight-inch manufacturing will influence the next stage of cost competition. Twelve-inch technology represents a longer-term frontier that could eventually reshape both power electronics and advanced computing.
Silicon carbide is not leaving the stage. It is simply entering the part of the technology cycle where manufacturing execution matters more than market excitement.
english
français
Deutsch
Italiano
Русский
Español
português
Nederlandse
ελληνικά
日本語
한국
العربية
Türkçe
polski